feat(secrets-mcp): 共享 key 删除时自动迁移并重定向 (v0.3.7)

删除仍被 metadata.key_ref 引用的 key 条目时，在同一事务内将密文复制到首个引用方，
其余引用方的 key_ref 重定向到新 owner；env_map 解析 key_ref 时不再限定 type=key。
Web 删除 API 返回 migrated；Dashboard 删除成功后提示迁移。

Bump secrets-mcp to 0.3.7；补充删除迁移相关单测（需 SECRETS_DATABASE_URL）。

Made-with: Cursor

AGENTS.md

@@ -29,7 +29,8 @@ secrets/
 
 - **建议库名**：`secrets-mcp`（专用实例，与历史库名区分）。
 - **连接**：环境变量 **`SECRETS_DATABASE_URL`**（本分支无本地配置文件路径）。
-- **表**：`entries`（含 `user_id`）、`secrets`、`entries_history`、`secrets_history`、`audit_log`、`users`、`oauth_accounts`，首次连接 **auto-migrate**。
+- **表**：`entries`（含 `user_id`）、`secrets`、`entries_history`、`secrets_history`、`audit_log`、`users`、`oauth_accounts`，首次连接 **auto-migrate**（`secrets-core` 的 `migrate`）。
+- **Web 会话**：与上项 **同一数据库 URL**；`secrets-mcp` 启动时对 tower-sessions 的 PostgreSQL 存储 **auto-migrate**（会话表与业务表共存于该实例，无需第二套连接串）。
 
 ### 表结构（摘录）
 
@@ -37,15 +38,18 @@ secrets/
 entries (
   id          UUID PRIMARY KEY DEFAULT uuidv7(),
   user_id     UUID,                                -- 多租户：NULL=遗留行；非空=归属用户
-  namespace   VARCHAR(64)   NOT NULL,
-  kind        VARCHAR(64)   NOT NULL,
+  folder      VARCHAR(128)  NOT NULL DEFAULT '',
+  type        VARCHAR(64)   NOT NULL DEFAULT '',
   name        VARCHAR(256)  NOT NULL,
+  notes       TEXT          NOT NULL DEFAULT '',
   tags        TEXT[]        NOT NULL DEFAULT '{}',
   metadata    JSONB         NOT NULL DEFAULT '{}',
   version     BIGINT        NOT NULL DEFAULT 1,
   created_at  TIMESTAMPTZ   NOT NULL DEFAULT NOW(),
   updated_at  TIMESTAMPTZ   NOT NULL DEFAULT NOW()
 )
+-- 唯一：UNIQUE(user_id, folder, name) WHERE user_id IS NOT NULL；
+--       UNIQUE(folder, name) WHERE user_id IS NULL（单租户遗留）
 ```
 
 ```sql
@@ -82,22 +86,31 @@ oauth_accounts (
   user_id      UUID NOT NULL REFERENCES users(id) ON DELETE CASCADE,
   provider     VARCHAR(32) NOT NULL,
   provider_id  VARCHAR(256) NOT NULL,
-  ...
+  email        VARCHAR(256),
+  name         VARCHAR(256),
+  avatar_url   TEXT,
+  created_at   TIMESTAMPTZ NOT NULL DEFAULT NOW(),
   UNIQUE(provider, provider_id)
 )
+-- 另有唯一索引 UNIQUE(user_id, provider)（迁移中 idx_oauth_accounts_user_provider）：同一用户每种 provider 至多一条关联。
 ```
 
 ### audit_log / history
 
-与迁移脚本一致：`audit_log`、`entries_history`、`secrets_history` 用于审计与时间旅行恢复；字段定义见 `crates/secrets-core/src/db.rs` 内 `migrate` SQL。`audit_log` 中普通业务事件的 `namespace/kind/name` 对应 entry 坐标；登录类事件固定使用 `namespace='auth'`，此时 `kind/name` 表示认证目标而非 entry 身份。
+与迁移脚本一致：`audit_log`、`entries_history`、`secrets_history` 用于审计与时间旅行恢复；字段定义见 `crates/secrets-core/src/db.rs` 内 `migrate` SQL。`audit_log` 含可选 **`user_id`**（多租户下标识操作者；可空以兼容遗留数据）。`audit_log` 中普通业务事件使用 **`folder` / `type` / `name`** 对应 entry 坐标；登录类事件固定使用 **`folder='auth'`**，此时 `type`/`name` 表示认证目标而非 entry 身份。
+
+### MCP 消歧（AI 调用）
+
+按 `name` 定位条目的工具（`get` / `update` / 单条 `delete` / `history` / `rollback`）：若该用户下仅一条匹配则直接执行；若多条（同 `name`、不同 `folder`）则返回错误并提示补全 `folder`。`secrets_delete` 的 `dry_run=true` 与真实删除使用相同消歧规则。
 
 ### 字段职责
 
 | 字段 | 含义 | 示例 |
 |------|------|------|
-| `namespace` | 隔离空间 | `refining` |
-| `kind` | 记录类型 | `server`, `service`, `key` |
-| `name` | 标识名 | `gitea`, `i-example0…` |
+| `folder` | 隔离空间（参与唯一键） | `refining` |
+| `type` | 软分类（不参与唯一键） | `server`, `service`, `key`, `person` |
+| `name` | 标识名 | `gitea`, `aliyun` |
+| `notes` | 非敏感说明 | 自由文本 |
 | `tags` | 标签 | `["aliyun","prod"]` |
 | `metadata` | 明文描述 | `ip`、`url`、`key_ref` |
 | `secrets.field_name` | 加密字段名（明文） | `token`, `ssh_key` |
@@ -105,7 +118,7 @@ oauth_accounts (
 
 ### PEM 共享（`key_ref`）
 
-将共享 PEM 存为 `kind=key` 的 entry；其它记录在 `metadata.key_ref` 指向该 key 的 `name`。更新 key 记录后，引用方通过服务层解析合并逻辑即可使用新密钥（实现见 `secrets_core::service`）。
+建议将共享 PEM 存为 **`type=key`** 的 entry；其它记录在 `metadata.key_ref` 指向目标 entry 的 `name`（支持 `folder/name` 格式消歧）。删除被引用 key 时，服务会自动迁移为单副本 + 重定向（复制到首个引用方，其余引用方改指向新 owner）；解析逻辑见 `secrets_core::service::env_map`。
 
 ## 代码规范
 

Cargo.lock

@@ -1968,7 +1968,7 @@ dependencies = [
 
 [[package]]
 name = "secrets-mcp"
-version = "0.2.2"
+version = "0.3.7"
 dependencies = [
  "anyhow",
  "askama",

README.md

@@ -17,7 +17,10 @@ cargo build --release -p secrets-mcp
 
 | 变量 | 说明 |
 |------|------|
-| `SECRETS_DATABASE_URL` | **必填**。PostgreSQL 连接串（建议专用库，如 `secrets-mcp`）。 |
+| `SECRETS_DATABASE_URL` | **必填**。PostgreSQL 连接串（推荐使用域名，例如 `db.refining.ltd`，避免直连 IP）。 |
+| `SECRETS_DATABASE_SSL_MODE` | 可选但强烈建议生产必填。推荐 `verify-full`（至少 `verify-ca`），避免回退到弱 TLS 模式。 |
+| `SECRETS_DATABASE_SSL_ROOT_CERT` | 可选。私有 CA 或自签链路时指定 CA 根证书路径（如 `/etc/secrets/pg-ca.crt`）。 |
+| `SECRETS_ENV` | 可选。设为 `prod` / `production` 时会拒绝弱 PostgreSQL TLS 模式（`prefer`、`disable`、`allow`、`require`）。 |
 | `BASE_URL` | 对外访问基址；OAuth 回调为 `{BASE_URL}/auth/google/callback`。默认 `http://localhost:9315`。 |
 | `SECRETS_MCP_BIND` | 监听地址，默认 `127.0.0.1:9315`。容器内或直接对外暴露端口时请改为 `0.0.0.0:9315`；反代时常为 `127.0.0.1:9315`。 |
 | `GOOGLE_CLIENT_ID` / `GOOGLE_CLIENT_SECRET` | 可选；不配置则无 Google 登录入口。运行时从环境读取，勿写入 CI、勿打入二进制。 |
@@ -27,8 +30,34 @@ cargo build --release -p secrets-mcp
 cargo run -p secrets-mcp
 ```
 
+生产推荐示例（PostgreSQL TLS）：
+
+```bash
+SECRETS_DATABASE_URL=postgres://postgres:***@db.refining.ltd:5432/secrets-mcp
+SECRETS_DATABASE_SSL_MODE=verify-full
+SECRETS_DATABASE_SSL_ROOT_CERT=/etc/secrets/pg-ca.crt
+SECRETS_ENV=production
+```
+
 - **Web**：`BASE_URL`（登录、Dashboard、设置密码短语、创建 API Key）。
-- **MCP**：Streamable HTTP 基址 `{BASE_URL}/mcp`，需 `Authorization: Bearer <api_key>` + `X-Encryption-Key: <hex>` 请求头。
+- **MCP**：Streamable HTTP 基址 `{BASE_URL}/mcp`，需 `Authorization: Bearer <api_key>` + `X-Encryption-Key: <hex>` 请求头（读密文工具须带密钥）。
+
+## PostgreSQL TLS 加固
+
+- 推荐将数据库域名单独设置为 `db.refining.ltd`，服务域名保持 `secrets.refining.app`。
+- 数据库证书建议使用可校验链路（如 Let's Encrypt 或私有 CA），并保证证书 `SAN` 包含 `db.refining.ltd`。
+- PostgreSQL 侧建议使用 `hostssl` 规则限制应用来源（如 `47.238.146.244/32`），逐步移除公网明文 `host` 访问。
+- 应用端推荐 `SECRETS_DATABASE_SSL_MODE=verify-full`；仅在过渡阶段可临时用 `verify-ca`。
+- 可执行运维步骤见 [`deploy/postgres-tls-hardening.md`](deploy/postgres-tls-hardening.md)。
+
+## MCP 与 AI 工作流（v0.3+）
+
+条目在逻辑上以 **`(folder, name)`** 在用户内唯一（数据库唯一索引：`user_id + folder + name`）。同名可在不同 folder 下各存一条（例如 `refining/aliyun` 与 `ricnsmart/aliyun`）。
+
+- **`secrets_search`**：发现条目（可按 query / folder / type / name 过滤）；不要求加密头。
+- **`secrets_get` / `secrets_update` / `secrets_delete`（按 name）/ `secrets_history` / `secrets_rollback`**：仅 `name` 且全局唯一则直接命中；若多条同名，返回消歧错误，需在参数中补 **`folder`**。
+- **`secrets_delete`**：`dry_run=true` 时与真实删除相同的消歧规则——唯一则预览一条，多条则报错并要求 `folder`。
+- **共享 key 自动迁移删除**：删除仍被 `metadata.key_ref` 引用的 key 条目时，系统会自动迁移：把密文复制到首个引用方，并将其余引用方的 `key_ref` 重定向到新 owner，然后继续删除。
 
 ## 加密架构（混合 E2EE）
 
@@ -122,13 +151,14 @@ flowchart LR
 
 ## 数据模型
 
-主表 **`entries`**（`namespace`、`kind`、`name`、`tags`、`metadata`，多租户时带 `user_id`）+ 子表 **`secrets`**（每行一个加密字段：`field_name`、`encrypted`）。另有 `entries_history`、`secrets_history`、`audit_log`，以及 **`users`**（含 `key_salt`、`key_check`、`key_params`、`api_key`）、**`oauth_accounts`**。首次连库自动迁移建表。
+主表 **`entries`**（`folder`、`type`、`name`、`notes`、`tags`、`metadata`，多租户时带 `user_id`）+ 子表 **`secrets`**（每行一个加密字段：`field_name`、`encrypted`）。**唯一性**：`UNIQUE(user_id, folder, name)`（`user_id` 为空时为遗留行唯一 `(folder, name)`）。另有 `entries_history`、`secrets_history`、`audit_log`，以及 **`users`**（含 `key_salt`、`key_check`、`key_params`、`api_key`）、**`oauth_accounts`**。首次连库自动迁移建表（`secrets-core` 的 `migrate`）；已有库可对照 [`scripts/migrate-v0.3.0.sql`](scripts/migrate-v0.3.0.sql) 做列重命名与索引重建。**Web 登录会话**（tower-sessions）使用同一 `SECRETS_DATABASE_URL`，进程启动时对会话存储执行迁移（见 `secrets-mcp` 中 `PostgresStore::migrate`），无需额外环境变量。
 
 | 位置 | 字段 | 说明 |
 |------|------|------|
-| entries | namespace | 一级隔离，如 `refining`、`ricnsmart` |
-| entries | kind | `server`、`service`、`key` 等（可扩展） |
-| entries | name | 人类可读标识 |
+| entries | folder | 组织/隔离空间，如 `refining`、`ricnsmart`；参与唯一键 |
+| entries | type | 软分类，如 `server`、`service`、`key`、`person`（可扩展，不参与唯一键） |
+| entries | name | 人类可读标识；与 `folder` 一起在用户内唯一 |
+| entries | notes | 非敏感说明文本 |
 | entries | metadata | 明文 JSON（ip、url、`key_ref` 等） |
 | secrets | field_name | 明文字段名，便于 schema 展示 |
 | secrets | encrypted | AES-GCM 密文（含 nonce） |
@@ -138,15 +168,16 @@ flowchart LR
 
 ### PEM 共享（`key_ref`）
 
-同一 PEM 可被多条 `server` 记录引用：将 PEM 存为 `kind=key` 的 entry，在服务器条目的 `metadata.key_ref` 中写 key 的名称；轮换时只更新 key 对应记录即可。
+同一 PEM 可被多条 `server` 等记录引用：建议将 PEM 存为 **`type=key`** 的 entry，在其它条目的 `metadata.key_ref` 中写目标 entry 的 `name`（支持 `folder/name` 格式消歧）；轮换时只更新该目标记录即可。
+删除共享 key 时，系统会自动迁移引用：将密文复制到首个引用方（单副本），其余引用方的 `key_ref` 自动重定向到该新 owner，再删除原 key 记录。
 
 ## 审计日志
 
-`add`、`update`、`delete` 等写操作写入 **`audit_log`**（操作类型、对象、摘要，不含 secret 明文）。
-其中业务条目事件使用 `[namespace/kind] name` 语义；登录类事件使用 `namespace='auth'`，此时 `kind/name` 表示认证目标（例如 `oauth/google`），不表示某条 secrets entry。
+`add`、`update`、`delete` 等写操作写入 **`audit_log`**（操作类型、对象、摘要，不含 secret 明文）。多租户场景下可写 **`user_id`**（可空，兼容遗留行）。
+业务条目事件使用 **`folder` / `type` / `name`**；登录类事件使用 **`folder='auth'`**，此时 `type`/`name` 表示认证目标（例如 `oauth` / `google`），不表示某条 secrets entry。
 
 ```sql
-SELECT action, namespace, kind, name, detail, created_at
+SELECT action, folder, type, name, detail, user_id, created_at
 FROM audit_log
 ORDER BY created_at DESC
 LIMIT 20;
@@ -159,6 +190,7 @@ Cargo.toml
 crates/secrets-core/     # db / crypto / models / audit / service
 crates/secrets-mcp/     # MCP HTTP、Web、OAuth、API Key
 scripts/
+  migrate-v0.3.0.sql    # 可选：手动 SQL 迁移（namespace/kind → folder/type、唯一键含 folder）
 deploy/                 # systemd、.env 示例
 ```
 

crates/secrets-core/src/audit.rs

@@ -3,7 +3,7 @@ use sqlx::{PgPool, Postgres, Transaction};
 use uuid::Uuid;
 
 pub const ACTION_LOGIN: &str = "login";
-pub const NAMESPACE_AUTH: &str = "auth";
+pub const FOLDER_AUTH: &str = "auth";
 
 fn login_detail(provider: &str, client_ip: Option<&str>, user_agent: Option<&str>) -> Value {
     json!({
@@ -16,7 +16,7 @@ fn login_detail(provider: &str, client_ip: Option<&str>, user_agent: Option<&str
 /// Write a login audit entry without requiring an explicit transaction.
 pub async fn log_login(
     pool: &PgPool,
-    kind: &str,
+    entry_type: &str,
     provider: &str,
     user_id: Uuid,
     client_ip: Option<&str>,
@@ -24,22 +24,22 @@ pub async fn log_login(
 ) {
     let detail = login_detail(provider, client_ip, user_agent);
     let result: Result<_, sqlx::Error> = sqlx::query(
-        "INSERT INTO audit_log (user_id, action, namespace, kind, name, detail) \
+        "INSERT INTO audit_log (user_id, action, folder, type, name, detail) \
          VALUES ($1, $2, $3, $4, $5, $6)",
     )
     .bind(user_id)
     .bind(ACTION_LOGIN)
-    .bind(NAMESPACE_AUTH)
-    .bind(kind)
+    .bind(FOLDER_AUTH)
+    .bind(entry_type)
     .bind(provider)
     .bind(&detail)
     .execute(pool)
     .await;
 
     if let Err(e) = result {
-        tracing::warn!(error = %e, kind, provider, "failed to write login audit log");
+        tracing::warn!(error = %e, entry_type, provider, "failed to write login audit log");
     } else {
-        tracing::debug!(kind, provider, ?user_id, "login audit logged");
+        tracing::debug!(entry_type, provider, ?user_id, "login audit logged");
     }
 }
 
@@ -48,19 +48,19 @@ pub async fn log_tx(
     tx: &mut Transaction<'_, Postgres>,
     user_id: Option<Uuid>,
     action: &str,
-    namespace: &str,
-    kind: &str,
+    folder: &str,
+    entry_type: &str,
     name: &str,
     detail: Value,
 ) {
     let result: Result<_, sqlx::Error> = sqlx::query(
-        "INSERT INTO audit_log (user_id, action, namespace, kind, name, detail) \
+        "INSERT INTO audit_log (user_id, action, folder, type, name, detail) \
          VALUES ($1, $2, $3, $4, $5, $6)",
     )
     .bind(user_id)
     .bind(action)
-    .bind(namespace)
-    .bind(kind)
+    .bind(folder)
+    .bind(entry_type)
     .bind(name)
     .bind(&detail)
     .execute(&mut **tx)
@@ -69,7 +69,7 @@ pub async fn log_tx(
     if let Err(e) = result {
         tracing::warn!(error = %e, "failed to write audit log");
     } else {
-        tracing::debug!(action, namespace, kind, name, "audit logged");
+        tracing::debug!(action, folder, entry_type, name, "audit logged");
     }
 }
 

crates/secrets-core/src/config.rs

@@ -1,4 +1,15 @@
-use anyhow::Result;
+use std::path::PathBuf;
+
+use anyhow::{Context, Result};
+use sqlx::postgres::PgSslMode;
+
+#[derive(Debug, Clone)]
+pub struct DatabaseConfig {
+    pub url: String,
+    pub ssl_mode: Option<PgSslMode>,
+    pub ssl_root_cert: Option<PathBuf>,
+    pub enforce_strict_tls: bool,
+}
 
 /// Resolve database URL from environment.
 /// Priority: `SECRETS_DATABASE_URL` env var → error.
@@ -18,3 +29,54 @@ pub fn resolve_db_url(override_url: &str) -> Result<String> {
          Example: SECRETS_DATABASE_URL=postgres://user:pass@host:port/dbname"
     )
 }
+
+fn env_var_non_empty(name: &str) -> Option<String> {
+    std::env::var(name)
+        .ok()
+        .filter(|value| !value.trim().is_empty())
+}
+
+fn parse_ssl_mode_from_env() -> Result<Option<PgSslMode>> {
+    let Some(mode) = env_var_non_empty("SECRETS_DATABASE_SSL_MODE") else {
+        return Ok(None);
+    };
+
+    let parsed = mode.parse::<PgSslMode>().with_context(|| {
+        format!(
+            "Invalid SECRETS_DATABASE_SSL_MODE='{mode}'. Use one of: disable, allow, prefer, require, verify-ca, verify-full."
+        )
+    })?;
+    Ok(Some(parsed))
+}
+
+fn resolve_ssl_root_cert_from_env() -> Result<Option<PathBuf>> {
+    let Some(path) = env_var_non_empty("SECRETS_DATABASE_SSL_ROOT_CERT") else {
+        return Ok(None);
+    };
+    let path = PathBuf::from(path);
+    if !path.exists() {
+        anyhow::bail!(
+            "SECRETS_DATABASE_SSL_ROOT_CERT points to a missing file: {}",
+            path.display()
+        );
+    }
+    Ok(Some(path))
+}
+
+fn is_production_env() -> bool {
+    matches!(
+        env_var_non_empty("SECRETS_ENV")
+            .as_deref()
+            .map(|value| value.to_ascii_lowercase()),
+        Some(value) if value == "prod" || value == "production"
+    )
+}
+
+pub fn resolve_db_config(override_url: &str) -> Result<DatabaseConfig> {
+    Ok(DatabaseConfig {
+        url: resolve_db_url(override_url)?,
+        ssl_mode: parse_ssl_mode_from_env()?,
+        ssl_root_cert: resolve_ssl_root_cert_from_env()?,
+        enforce_strict_tls: is_production_env(),
+    })
+}

crates/secrets-core/src/db.rs

@@ -1,14 +1,45 @@
-use anyhow::Result;
+use std::str::FromStr;
+
+use anyhow::{Context, Result};
 use serde_json::Value;
 use sqlx::PgPool;
-use sqlx::postgres::PgPoolOptions;
+use sqlx::postgres::{PgConnectOptions, PgPoolOptions, PgSslMode};
 
-pub async fn create_pool(database_url: &str) -> Result<PgPool> {
+use crate::config::DatabaseConfig;
+
+fn build_connect_options(config: &DatabaseConfig) -> Result<PgConnectOptions> {
+    let mut options = PgConnectOptions::from_str(&config.url)
+        .with_context(|| "failed to parse SECRETS_DATABASE_URL".to_string())?;
+
+    if let Some(mode) = config.ssl_mode {
+        options = options.ssl_mode(mode);
+    }
+    if let Some(path) = &config.ssl_root_cert {
+        options = options.ssl_root_cert(path);
+    }
+
+    if config.enforce_strict_tls
+        && !matches!(
+            options.get_ssl_mode(),
+            PgSslMode::VerifyCa | PgSslMode::VerifyFull
+        )
+    {
+        anyhow::bail!(
+            "Refusing to start in production with weak PostgreSQL TLS mode. \
+             Set SECRETS_DATABASE_SSL_MODE=verify-ca or verify-full."
+        );
+    }
+
+    Ok(options)
+}
+
+pub async fn create_pool(config: &DatabaseConfig) -> Result<PgPool> {
     tracing::debug!("connecting to database");
+    let connect_options = build_connect_options(config)?;
     let pool = PgPoolOptions::new()
         .max_connections(10)
         .acquire_timeout(std::time::Duration::from_secs(5))
-        .connect(database_url)
+        .connect_with(connect_options)
         .await?;
     tracing::debug!("database connection established");
     Ok(pool)
@@ -22,9 +53,10 @@ pub async fn migrate(pool: &PgPool) -> Result<()> {
         CREATE TABLE IF NOT EXISTS entries (
             id          UUID PRIMARY KEY DEFAULT uuidv7(),
             user_id     UUID,
-            namespace   VARCHAR(64)  NOT NULL,
-            kind        VARCHAR(64)  NOT NULL,
+            folder      VARCHAR(128) NOT NULL DEFAULT '',
+            type        VARCHAR(64)  NOT NULL DEFAULT '',
             name        VARCHAR(256) NOT NULL,
+            notes       TEXT         NOT NULL DEFAULT '',
             tags        TEXT[]       NOT NULL DEFAULT '{}',
             metadata    JSONB        NOT NULL DEFAULT '{}',
             version     BIGINT       NOT NULL DEFAULT 1,
@@ -34,19 +66,19 @@ pub async fn migrate(pool: &PgPool) -> Result<()> {
 
         -- Legacy unique constraint without user_id (single-user mode)
         CREATE UNIQUE INDEX IF NOT EXISTS idx_entries_unique_legacy
-            ON entries(namespace, kind, name)
+            ON entries(folder, name)
             WHERE user_id IS NULL;
 
         -- Multi-user unique constraint
         CREATE UNIQUE INDEX IF NOT EXISTS idx_entries_unique_user
-            ON entries(user_id, namespace, kind, name)
+            ON entries(user_id, folder, name)
             WHERE user_id IS NOT NULL;
 
-        CREATE INDEX IF NOT EXISTS idx_entries_namespace ON entries(namespace);
-        CREATE INDEX IF NOT EXISTS idx_entries_kind      ON entries(kind);
-        CREATE INDEX IF NOT EXISTS idx_entries_user_id   ON entries(user_id) WHERE user_id IS NOT NULL;
-        CREATE INDEX IF NOT EXISTS idx_entries_tags      ON entries USING GIN(tags);
-        CREATE INDEX IF NOT EXISTS idx_entries_metadata  ON entries USING GIN(metadata jsonb_path_ops);
+        CREATE INDEX IF NOT EXISTS idx_entries_folder   ON entries(folder) WHERE folder <> '';
+        CREATE INDEX IF NOT EXISTS idx_entries_type     ON entries(type) WHERE type <> '';
+        CREATE INDEX IF NOT EXISTS idx_entries_user_id  ON entries(user_id) WHERE user_id IS NOT NULL;
+        CREATE INDEX IF NOT EXISTS idx_entries_tags     ON entries USING GIN(tags);
+        CREATE INDEX IF NOT EXISTS idx_entries_metadata ON entries USING GIN(metadata jsonb_path_ops);
 
         -- ── secrets: one row per encrypted field ─────────────────────────────────
         CREATE TABLE IF NOT EXISTS secrets (
@@ -67,23 +99,23 @@ pub async fn migrate(pool: &PgPool) -> Result<()> {
             id          BIGINT GENERATED ALWAYS AS IDENTITY PRIMARY KEY,
             user_id     UUID,
             action      VARCHAR(32)  NOT NULL,
-            namespace   VARCHAR(64)  NOT NULL,
-            kind        VARCHAR(64)  NOT NULL,
+            folder      VARCHAR(128) NOT NULL DEFAULT '',
+            type        VARCHAR(64)  NOT NULL DEFAULT '',
             name        VARCHAR(256) NOT NULL,
             detail      JSONB        NOT NULL DEFAULT '{}',
             created_at  TIMESTAMPTZ  NOT NULL DEFAULT NOW()
         );
 
-        CREATE INDEX IF NOT EXISTS idx_audit_log_created  ON audit_log(created_at DESC);
-        CREATE INDEX IF NOT EXISTS idx_audit_log_ns_kind  ON audit_log(namespace, kind);
-        CREATE INDEX IF NOT EXISTS idx_audit_log_user_id  ON audit_log(user_id) WHERE user_id IS NOT NULL;
+        CREATE INDEX IF NOT EXISTS idx_audit_log_created      ON audit_log(created_at DESC);
+        CREATE INDEX IF NOT EXISTS idx_audit_log_folder_type  ON audit_log(folder, type);
+        CREATE INDEX IF NOT EXISTS idx_audit_log_user_id      ON audit_log(user_id) WHERE user_id IS NOT NULL;
 
         -- ── entries_history ───────────────────────────────────────────────────────
         CREATE TABLE IF NOT EXISTS entries_history (
             id          BIGINT GENERATED ALWAYS AS IDENTITY PRIMARY KEY,
             entry_id    UUID         NOT NULL,
-            namespace   VARCHAR(64)  NOT NULL,
-            kind        VARCHAR(64)  NOT NULL,
+            folder      VARCHAR(128) NOT NULL DEFAULT '',
+            type        VARCHAR(64)  NOT NULL DEFAULT '',
             name        VARCHAR(256) NOT NULL,
             version     BIGINT       NOT NULL,
             action      VARCHAR(16)  NOT NULL,
@@ -94,8 +126,8 @@ pub async fn migrate(pool: &PgPool) -> Result<()> {
 
         CREATE INDEX IF NOT EXISTS idx_entries_history_entry_id
             ON entries_history(entry_id, version DESC);
-        CREATE INDEX IF NOT EXISTS idx_entries_history_ns_kind_name
-            ON entries_history(namespace, kind, name, version DESC);
+        CREATE INDEX IF NOT EXISTS idx_entries_history_folder_type_name
+            ON entries_history(folder, type, name, version DESC);
 
         -- Backfill: add user_id to entries_history for multi-tenant isolation
         ALTER TABLE entries_history ADD COLUMN IF NOT EXISTS user_id UUID;
@@ -103,6 +135,9 @@ pub async fn migrate(pool: &PgPool) -> Result<()> {
             ON entries_history(user_id) WHERE user_id IS NOT NULL;
         ALTER TABLE entries_history DROP COLUMN IF EXISTS actor;
 
+        -- Backfill: add notes to entries if not present (fresh installs already have it)
+        ALTER TABLE entries ADD COLUMN IF NOT EXISTS notes TEXT NOT NULL DEFAULT '';
+
         -- ── secrets_history: field-level snapshot ────────────────────────────────
         CREATE TABLE IF NOT EXISTS secrets_history (
             id              BIGINT GENERATED ALWAYS AS IDENTITY PRIMARY KEY,
@@ -123,9 +158,6 @@ pub async fn migrate(pool: &PgPool) -> Result<()> {
         -- Drop redundant actor column (derivable via entries_history JOIN)
         ALTER TABLE secrets_history DROP COLUMN IF EXISTS actor;
 
-        -- Drop redundant actor column; user_id already identifies the business user
-        ALTER TABLE audit_log DROP COLUMN IF EXISTS actor;
-
         -- ── users ─────────────────────────────────────────────────────────────────
         CREATE TABLE IF NOT EXISTS users (
             id          UUID PRIMARY KEY DEFAULT uuidv7(),
@@ -191,12 +223,179 @@ pub async fn migrate(pool: &PgPool) -> Result<()> {
     )
     .execute(pool)
     .await?;
+    migrate_schema(pool).await?;
     restore_plaintext_api_keys(pool).await?;
 
     tracing::debug!("migrations complete");
     Ok(())
 }
 
+/// Idempotent schema migration: rename namespace→folder, kind→type in existing databases.
+async fn migrate_schema(pool: &PgPool) -> Result<()> {
+    sqlx::raw_sql(
+        r#"
+        -- ── entries: rename namespace→folder, kind→type ──────────────────────────
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'entries' AND column_name = 'namespace'
+            ) THEN
+                ALTER TABLE entries RENAME COLUMN namespace TO folder;
+            END IF;
+        END $$;
+
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'entries' AND column_name = 'kind'
+            ) THEN
+                ALTER TABLE entries RENAME COLUMN kind TO type;
+            END IF;
+        END $$;
+
+        -- ── audit_log: rename namespace→folder, kind→type ────────────────────────
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'audit_log' AND column_name = 'namespace'
+            ) THEN
+                ALTER TABLE audit_log RENAME COLUMN namespace TO folder;
+            END IF;
+        END $$;
+
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'audit_log' AND column_name = 'kind'
+            ) THEN
+                ALTER TABLE audit_log RENAME COLUMN kind TO type;
+            END IF;
+        END $$;
+
+        -- ── entries_history: rename namespace→folder, kind→type ──────────────────
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'entries_history' AND column_name = 'namespace'
+            ) THEN
+                ALTER TABLE entries_history RENAME COLUMN namespace TO folder;
+            END IF;
+        END $$;
+
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'entries_history' AND column_name = 'kind'
+            ) THEN
+                ALTER TABLE entries_history RENAME COLUMN kind TO type;
+            END IF;
+        END $$;
+
+        -- ── Set empty defaults for new folder/type columns ────────────────────────
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'entries' AND column_name = 'folder'
+            ) THEN
+                UPDATE entries SET folder = '' WHERE folder IS NULL;
+                ALTER TABLE entries ALTER COLUMN folder SET NOT NULL;
+                ALTER TABLE entries ALTER COLUMN folder SET DEFAULT '';
+            END IF;
+        END $$;
+
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'entries' AND column_name = 'type'
+            ) THEN
+                UPDATE entries SET type = '' WHERE type IS NULL;
+                ALTER TABLE entries ALTER COLUMN type SET NOT NULL;
+                ALTER TABLE entries ALTER COLUMN type SET DEFAULT '';
+            END IF;
+        END $$;
+
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'audit_log' AND column_name = 'folder'
+            ) THEN
+                UPDATE audit_log SET folder = '' WHERE folder IS NULL;
+                ALTER TABLE audit_log ALTER COLUMN folder SET NOT NULL;
+                ALTER TABLE audit_log ALTER COLUMN folder SET DEFAULT '';
+            END IF;
+        END $$;
+
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'audit_log' AND column_name = 'type'
+            ) THEN
+                UPDATE audit_log SET type = '' WHERE type IS NULL;
+                ALTER TABLE audit_log ALTER COLUMN type SET NOT NULL;
+                ALTER TABLE audit_log ALTER COLUMN type SET DEFAULT '';
+            END IF;
+        END $$;
+
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'entries_history' AND column_name = 'folder'
+            ) THEN
+                UPDATE entries_history SET folder = '' WHERE folder IS NULL;
+                ALTER TABLE entries_history ALTER COLUMN folder SET NOT NULL;
+                ALTER TABLE entries_history ALTER COLUMN folder SET DEFAULT '';
+            END IF;
+        END $$;
+
+        DO $$ BEGIN
+            IF EXISTS (
+                SELECT 1 FROM information_schema.columns
+                WHERE table_name = 'entries_history' AND column_name = 'type'
+            ) THEN
+                UPDATE entries_history SET type = '' WHERE type IS NULL;
+                ALTER TABLE entries_history ALTER COLUMN type SET NOT NULL;
+                ALTER TABLE entries_history ALTER COLUMN type SET DEFAULT '';
+            END IF;
+        END $$;
+
+        -- ── Rebuild unique indexes on entries: folder is now part of the key ────────
+        -- (user_id, folder, name) allows same name in different folders.
+        DROP INDEX IF EXISTS idx_entries_unique_legacy;
+        DROP INDEX IF EXISTS idx_entries_unique_user;
+
+        CREATE UNIQUE INDEX IF NOT EXISTS idx_entries_unique_legacy
+            ON entries(folder, name)
+            WHERE user_id IS NULL;
+
+        CREATE UNIQUE INDEX IF NOT EXISTS idx_entries_unique_user
+            ON entries(user_id, folder, name)
+            WHERE user_id IS NOT NULL;
+
+        -- ── Replace old namespace/kind indexes ────────────────────────────────────
+        DROP INDEX IF EXISTS idx_entries_namespace;
+        DROP INDEX IF EXISTS idx_entries_kind;
+        DROP INDEX IF EXISTS idx_audit_log_ns_kind;
+        DROP INDEX IF EXISTS idx_entries_history_ns_kind_name;
+
+        CREATE INDEX IF NOT EXISTS idx_entries_folder
+            ON entries(folder) WHERE folder <> '';
+        CREATE INDEX IF NOT EXISTS idx_entries_type
+            ON entries(type) WHERE type <> '';
+        CREATE INDEX IF NOT EXISTS idx_audit_log_folder_type
+            ON audit_log(folder, type);
+        CREATE INDEX IF NOT EXISTS idx_entries_history_folder_type_name
+            ON entries_history(folder, type, name, version DESC);
+
+        -- ── Drop legacy actor columns ─────────────────────────────────────────────
+        ALTER TABLE secrets_history DROP COLUMN IF EXISTS actor;
+        ALTER TABLE audit_log DROP COLUMN IF EXISTS actor;
+        "#,
+    )
+    .execute(pool)
+    .await?;
+    Ok(())
+}
+
 async fn restore_plaintext_api_keys(pool: &PgPool) -> Result<()> {
     let has_users_api_key: bool = sqlx::query_scalar(
         "SELECT EXISTS (
@@ -265,8 +464,8 @@ async fn restore_plaintext_api_keys(pool: &PgPool) -> Result<()> {
 pub struct EntrySnapshotParams<'a> {
     pub entry_id: uuid::Uuid,
     pub user_id: Option<uuid::Uuid>,
-    pub namespace: &'a str,
-    pub kind: &'a str,
+    pub folder: &'a str,
+    pub entry_type: &'a str,
     pub name: &'a str,
     pub version: i64,
     pub action: &'a str,
@@ -280,12 +479,12 @@ pub async fn snapshot_entry_history(
 ) -> Result<()> {
     sqlx::query(
         "INSERT INTO entries_history \
-         (entry_id, namespace, kind, name, version, action, tags, metadata, user_id) \
+         (entry_id, folder, type, name, version, action, tags, metadata, user_id) \
          VALUES ($1, $2, $3, $4, $5, $6, $7, $8, $9)",
     )
     .bind(p.entry_id)
-    .bind(p.namespace)
-    .bind(p.kind)
+    .bind(p.folder)
+    .bind(p.entry_type)
     .bind(p.name)
     .bind(p.version)
     .bind(p.action)

crates/secrets-core/src/models.rs

@@ -4,15 +4,18 @@ use serde_json::Value;
 use std::collections::BTreeMap;
 use uuid::Uuid;
 
-/// A top-level entry (server, service, key, …).
+/// A top-level entry (server, service, key, person, …).
 /// Sensitive fields are stored separately in `secrets`.
 #[derive(Debug, Serialize, Deserialize, sqlx::FromRow)]
 pub struct Entry {
     pub id: Uuid,
     pub user_id: Option<Uuid>,
-    pub namespace: String,
-    pub kind: String,
+    pub folder: String,
+    #[serde(rename = "type")]
+    #[sqlx(rename = "type")]
+    pub entry_type: String,
     pub name: String,
+    pub notes: String,
     pub tags: Vec<String>,
     pub metadata: Value,
     pub version: i64,
@@ -40,8 +43,40 @@ pub struct SecretField {
 pub struct EntryRow {
     pub id: Uuid,
     pub version: i64,
+    pub folder: String,
+    #[sqlx(rename = "type")]
+    pub entry_type: String,
     pub tags: Vec<String>,
     pub metadata: Value,
+    pub notes: String,
+}
+
+/// Entry row including `name` (used for id-scoped web / service updates).
+#[derive(Debug, sqlx::FromRow)]
+pub struct EntryWriteRow {
+    pub id: Uuid,
+    pub version: i64,
+    pub folder: String,
+    #[sqlx(rename = "type")]
+    pub entry_type: String,
+    pub name: String,
+    pub tags: Vec<String>,
+    pub metadata: Value,
+    pub notes: String,
+}
+
+impl From<&EntryWriteRow> for EntryRow {
+    fn from(r: &EntryWriteRow) -> Self {
+        EntryRow {
+            id: r.id,
+            version: r.version,
+            folder: r.folder.clone(),
+            entry_type: r.entry_type.clone(),
+            tags: r.tags.clone(),
+            metadata: r.metadata.clone(),
+            notes: r.notes.clone(),
+        }
+    }
 }
 
 /// Minimal secret field row fetched before snapshots or cascade deletes.
@@ -128,10 +163,14 @@ pub struct ExportData {
 /// A single entry with decrypted secrets for export/import.
 #[derive(Debug, Serialize, Deserialize)]
 pub struct ExportEntry {
-    pub namespace: String,
-    pub kind: String,
     pub name: String,
     #[serde(default)]
+    pub folder: String,
+    #[serde(default, rename = "type")]
+    pub entry_type: String,
+    #[serde(default)]
+    pub notes: String,
+    #[serde(default)]
     pub tags: Vec<String>,
     #[serde(default)]
     pub metadata: Value,
@@ -181,8 +220,10 @@ pub struct AuditLogEntry {
     pub id: i64,
     pub user_id: Option<Uuid>,
     pub action: String,
-    pub namespace: String,
-    pub kind: String,
+    pub folder: String,
+    #[serde(rename = "type")]
+    #[sqlx(rename = "type")]
+    pub entry_type: String,
     pub name: String,
     pub detail: Value,
     pub created_at: DateTime<Utc>,

crates/secrets-core/src/service/add.rs

@@ -159,18 +159,20 @@ pub fn flatten_json_fields(prefix: &str, value: &Value) -> Vec<(String, Value)>
 
 #[derive(Debug, serde::Serialize)]
 pub struct AddResult {
-    pub namespace: String,
-    pub kind: String,
     pub name: String,
+    pub folder: String,
+    #[serde(rename = "type")]
+    pub entry_type: String,
     pub tags: Vec<String>,
     pub meta_keys: Vec<String>,
     pub secret_keys: Vec<String>,
 }
 
 pub struct AddParams<'a> {
-    pub namespace: &'a str,
-    pub kind: &'a str,
     pub name: &'a str,
+    pub folder: &'a str,
+    pub entry_type: &'a str,
+    pub notes: &'a str,
     pub tags: &'a [String],
     pub meta_entries: &'a [String],
     pub secret_entries: &'a [String],
@@ -186,25 +188,23 @@ pub async fn run(pool: &PgPool, params: AddParams<'_>, master_key: &[u8; 32]) ->
 
     let mut tx = pool.begin().await?;
 
-    // Fetch existing entry (user-scoped or global depending on user_id)
+    // Fetch existing entry by (user_id, folder, name) — the natural unique key
     let existing: Option<EntryRow> = if let Some(uid) = params.user_id {
         sqlx::query_as(
-            "SELECT id, version, tags, metadata FROM entries \
-             WHERE user_id = $1 AND namespace = $2 AND kind = $3 AND name = $4",
+            "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+             WHERE user_id = $1 AND folder = $2 AND name = $3",
         )
         .bind(uid)
-        .bind(params.namespace)
-        .bind(params.kind)
+        .bind(params.folder)
         .bind(params.name)
         .fetch_optional(&mut *tx)
         .await?
     } else {
         sqlx::query_as(
-            "SELECT id, version, tags, metadata FROM entries \
-             WHERE user_id IS NULL AND namespace = $1 AND kind = $2 AND name = $3",
+            "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+             WHERE user_id IS NULL AND folder = $1 AND name = $2",
         )
-        .bind(params.namespace)
-        .bind(params.kind)
+        .bind(params.folder)
         .bind(params.name)
         .fetch_optional(&mut *tx)
         .await?
@@ -216,8 +216,8 @@ pub async fn run(pool: &PgPool, params: AddParams<'_>, master_key: &[u8; 32]) ->
             db::EntrySnapshotParams {
                 entry_id: ex.id,
                 user_id: params.user_id,
-                namespace: params.namespace,
-                kind: params.kind,
+                folder: params.folder,
+                entry_type: params.entry_type,
                 name: params.name,
                 version: ex.version,
                 action: "add",
@@ -232,10 +232,13 @@ pub async fn run(pool: &PgPool, params: AddParams<'_>, master_key: &[u8; 32]) ->
 
     let entry_id: Uuid = if let Some(uid) = params.user_id {
         sqlx::query_scalar(
-            r#"INSERT INTO entries (user_id, namespace, kind, name, tags, metadata, version, updated_at)
-               VALUES ($1, $2, $3, $4, $5, $6, 1, NOW())
-               ON CONFLICT (user_id, namespace, kind, name) WHERE user_id IS NOT NULL
+            r#"INSERT INTO entries (user_id, folder, type, name, notes, tags, metadata, version, updated_at)
+               VALUES ($1, $2, $3, $4, $5, $6, $7, 1, NOW())
+               ON CONFLICT (user_id, folder, name) WHERE user_id IS NOT NULL
                DO UPDATE SET
+                   folder     = EXCLUDED.folder,
+                   type       = EXCLUDED.type,
+                   notes      = EXCLUDED.notes,
                    tags       = EXCLUDED.tags,
                    metadata   = EXCLUDED.metadata,
                    version    = entries.version + 1,
@@ -243,28 +246,33 @@ pub async fn run(pool: &PgPool, params: AddParams<'_>, master_key: &[u8; 32]) ->
                RETURNING id"#,
         )
         .bind(uid)
-        .bind(params.namespace)
-        .bind(params.kind)
+        .bind(params.folder)
+        .bind(params.entry_type)
         .bind(params.name)
+        .bind(params.notes)
         .bind(params.tags)
         .bind(&metadata)
         .fetch_one(&mut *tx)
         .await?
     } else {
         sqlx::query_scalar(
-            r#"INSERT INTO entries (namespace, kind, name, tags, metadata, version, updated_at)
-               VALUES ($1, $2, $3, $4, $5, 1, NOW())
-               ON CONFLICT (namespace, kind, name) WHERE user_id IS NULL
+            r#"INSERT INTO entries (folder, type, name, notes, tags, metadata, version, updated_at)
+               VALUES ($1, $2, $3, $4, $5, $6, 1, NOW())
+               ON CONFLICT (folder, name) WHERE user_id IS NULL
                DO UPDATE SET
+                   folder     = EXCLUDED.folder,
+                   type       = EXCLUDED.type,
+                   notes      = EXCLUDED.notes,
                    tags       = EXCLUDED.tags,
                    metadata   = EXCLUDED.metadata,
                    version    = entries.version + 1,
                    updated_at = NOW()
                RETURNING id"#,
         )
-        .bind(params.namespace)
-        .bind(params.kind)
+        .bind(params.folder)
+        .bind(params.entry_type)
         .bind(params.name)
+        .bind(params.notes)
         .bind(params.tags)
         .bind(&metadata)
         .fetch_one(&mut *tx)
@@ -282,8 +290,8 @@ pub async fn run(pool: &PgPool, params: AddParams<'_>, master_key: &[u8; 32]) ->
             db::EntrySnapshotParams {
                 entry_id,
                 user_id: params.user_id,
-                namespace: params.namespace,
-                kind: params.kind,
+                folder: params.folder,
+                entry_type: params.entry_type,
                 name: params.name,
                 version: new_entry_version,
                 action: "create",
@@ -348,8 +356,8 @@ pub async fn run(pool: &PgPool, params: AddParams<'_>, master_key: &[u8; 32]) ->
         &mut tx,
         params.user_id,
         "add",
-        params.namespace,
-        params.kind,
+        params.folder,
+        params.entry_type,
         params.name,
         serde_json::json!({
             "tags": params.tags,
@@ -362,9 +370,9 @@ pub async fn run(pool: &PgPool, params: AddParams<'_>, master_key: &[u8; 32]) ->
     tx.commit().await?;
 
     Ok(AddResult {
-        namespace: params.namespace.to_string(),
-        kind: params.kind.to_string(),
         name: params.name.to_string(),
+        folder: params.folder.to_string(),
+        entry_type: params.entry_type.to_string(),
         tags: params.tags.to_vec(),
         meta_keys,
         secret_keys,

crates/secrets-core/src/service/audit_log.rs

@@ -8,7 +8,7 @@ pub async fn list_for_user(pool: &PgPool, user_id: Uuid, limit: i64) -> Result<V
     let limit = limit.clamp(1, 200);
 
     let rows = sqlx::query_as(
-        "SELECT id, user_id, action, namespace, kind, name, detail, created_at \
+        "SELECT id, user_id, action, folder, type, name, detail, created_at \
          FROM audit_log \
          WHERE user_id = $1 \
          ORDER BY created_at DESC, id DESC \

crates/secrets-core/src/service/delete.rs

@@ -4,50 +4,274 @@ use sqlx::PgPool;
 use uuid::Uuid;
 
 use crate::db;
-use crate::models::{EntryRow, SecretFieldRow};
+use crate::models::{EntryRow, EntryWriteRow, SecretFieldRow};
 
 #[derive(Debug, serde::Serialize)]
 pub struct DeletedEntry {
-    pub namespace: String,
-    pub kind: String,
     pub name: String,
+    pub folder: String,
+    #[serde(rename = "type")]
+    pub entry_type: String,
 }
 
 #[derive(Debug, serde::Serialize)]
 pub struct DeleteResult {
     pub deleted: Vec<DeletedEntry>,
+    pub migrated: Vec<String>,
     pub dry_run: bool,
 }
 
 pub struct DeleteParams<'a> {
-    pub namespace: &'a str,
-    pub kind: Option<&'a str>,
+    /// If set, delete a single entry by name.
     pub name: Option<&'a str>,
+    /// Folder filter for bulk delete.
+    pub folder: Option<&'a str>,
+    /// Type filter for bulk delete.
+    pub entry_type: Option<&'a str>,
     pub dry_run: bool,
     pub user_id: Option<Uuid>,
 }
 
+#[derive(Debug, sqlx::FromRow)]
+struct KeyReferrer {
+    id: Uuid,
+    folder: String,
+    #[sqlx(rename = "type")]
+    entry_type: String,
+    name: String,
+}
+
+fn ref_label(r: &KeyReferrer) -> String {
+    format!("{}/{} ({})", r.folder, r.name, r.entry_type)
+}
+
+fn ref_path(r: &KeyReferrer) -> String {
+    format!("{}/{}", r.folder, r.name)
+}
+
+async fn fetch_key_referrers_pool(
+    pool: &PgPool,
+    key_entry_id: Uuid,
+    key_folder: &str,
+    key_name: &str,
+    user_id: Option<Uuid>,
+) -> Result<Vec<KeyReferrer>> {
+    let qualified = format!("{}/{}", key_folder, key_name);
+    let refs: Vec<KeyReferrer> = if let Some(uid) = user_id {
+        sqlx::query_as(
+            "SELECT id, folder, type, name FROM entries \
+             WHERE user_id = $1 AND id <> $2 \
+               AND (metadata->>'key_ref' = $3 OR metadata->>'key_ref' = $4) \
+             ORDER BY folder, type, name",
+        )
+        .bind(uid)
+        .bind(key_entry_id)
+        .bind(key_name)
+        .bind(&qualified)
+        .fetch_all(pool)
+        .await?
+    } else {
+        sqlx::query_as(
+            "SELECT id, folder, type, name FROM entries \
+             WHERE user_id IS NULL AND id <> $1 \
+               AND (metadata->>'key_ref' = $2 OR metadata->>'key_ref' = $3) \
+             ORDER BY folder, type, name",
+        )
+        .bind(key_entry_id)
+        .bind(key_name)
+        .bind(&qualified)
+        .fetch_all(pool)
+        .await?
+    };
+    Ok(refs)
+}
+
+async fn migrate_key_refs_if_needed(
+    tx: &mut sqlx::Transaction<'_, sqlx::Postgres>,
+    key_row: &EntryRow,
+    key_name: &str,
+    user_id: Option<Uuid>,
+    dry_run: bool,
+) -> Result<Vec<String>> {
+    let qualified = format!("{}/{}", key_row.folder, key_name);
+    let refs: Vec<KeyReferrer> = if let Some(uid) = user_id {
+        sqlx::query_as(
+            "SELECT id, folder, type, name FROM entries \
+             WHERE user_id = $1 AND id <> $2 \
+               AND (metadata->>'key_ref' = $3 OR metadata->>'key_ref' = $4) \
+             ORDER BY folder, type, name",
+        )
+        .bind(uid)
+        .bind(key_row.id)
+        .bind(key_name)
+        .bind(&qualified)
+        .fetch_all(&mut **tx)
+        .await?
+    } else {
+        sqlx::query_as(
+            "SELECT id, folder, type, name FROM entries \
+             WHERE user_id IS NULL AND id <> $1 \
+               AND (metadata->>'key_ref' = $2 OR metadata->>'key_ref' = $3) \
+             ORDER BY folder, type, name",
+        )
+        .bind(key_row.id)
+        .bind(key_name)
+        .bind(&qualified)
+        .fetch_all(&mut **tx)
+        .await?
+    };
+
+    if refs.is_empty() {
+        return Ok(vec![]);
+    }
+    if dry_run {
+        return Ok(refs.iter().map(ref_label).collect());
+    }
+
+    let owner = &refs[0];
+    let owner_path = ref_path(owner);
+    let key_fields: Vec<SecretFieldRow> =
+        sqlx::query_as("SELECT id, field_name, encrypted FROM secrets WHERE entry_id = $1")
+            .bind(key_row.id)
+            .fetch_all(&mut **tx)
+            .await?;
+
+    for f in &key_fields {
+        sqlx::query(
+            "INSERT INTO secrets (entry_id, field_name, encrypted) VALUES ($1, $2, $3) \
+             ON CONFLICT (entry_id, field_name) DO NOTHING",
+        )
+        .bind(owner.id)
+        .bind(&f.field_name)
+        .bind(&f.encrypted)
+        .execute(&mut **tx)
+        .await?;
+    }
+
+    sqlx::query(
+        "UPDATE entries SET metadata = metadata - 'key_ref', \
+         version = version + 1, updated_at = NOW() WHERE id = $1",
+    )
+    .bind(owner.id)
+    .execute(&mut **tx)
+    .await?;
+
+    crate::audit::log_tx(
+        tx,
+        user_id,
+        "key_migrate",
+        &owner.folder,
+        &owner.entry_type,
+        &owner.name,
+        json!({
+            "from_key": format!("{}/{}", key_row.folder, key_name),
+            "role": "new_owner",
+            "redirect_target": owner_path,
+        }),
+    )
+    .await;
+
+    for r in refs.iter().skip(1) {
+        sqlx::query(
+            "UPDATE entries SET metadata = jsonb_set(metadata, '{key_ref}', to_jsonb($2::text), true), \
+             version = version + 1, updated_at = NOW() WHERE id = $1",
+        )
+        .bind(r.id)
+        .bind(&owner_path)
+        .execute(&mut **tx)
+        .await?;
+
+        crate::audit::log_tx(
+            tx,
+            user_id,
+            "key_migrate",
+            &r.folder,
+            &r.entry_type,
+            &r.name,
+            json!({
+                "from_key": format!("{}/{}", key_row.folder, key_name),
+                "role": "redirected_ref",
+                "redirect_to": owner_path,
+            }),
+        )
+        .await;
+    }
+
+    Ok(refs.iter().map(ref_label).collect())
+}
+
+/// Delete a single entry by id (multi-tenant: `user_id` must match). Cascades `secrets` via FK.
+pub async fn delete_by_id(pool: &PgPool, entry_id: Uuid, user_id: Uuid) -> Result<DeleteResult> {
+    let mut tx = pool.begin().await?;
+    let row: Option<EntryWriteRow> = sqlx::query_as(
+        "SELECT id, version, folder, type, name, tags, metadata, notes FROM entries \
+         WHERE id = $1 AND user_id = $2 FOR UPDATE",
+    )
+    .bind(entry_id)
+    .bind(user_id)
+    .fetch_optional(&mut *tx)
+    .await?;
+
+    let row = match row {
+        Some(r) => r,
+        None => {
+            tx.rollback().await?;
+            anyhow::bail!("Entry not found");
+        }
+    };
+
+    let folder = row.folder.clone();
+    let entry_type = row.entry_type.clone();
+    let name = row.name.clone();
+    let entry_row: EntryRow = (&row).into();
+    let migrated =
+        migrate_key_refs_if_needed(&mut tx, &entry_row, &name, Some(user_id), false).await?;
+
+    snapshot_and_delete(
+        &mut tx,
+        &folder,
+        &entry_type,
+        &name,
+        &entry_row,
+        Some(user_id),
+    )
+    .await?;
+    crate::audit::log_tx(
+        &mut tx,
+        Some(user_id),
+        "delete",
+        &folder,
+        &entry_type,
+        &name,
+        json!({ "source": "web", "entry_id": entry_id }),
+    )
+    .await;
+    tx.commit().await?;
+
+    Ok(DeleteResult {
+        deleted: vec![DeletedEntry {
+            name,
+            folder,
+            entry_type,
+        }],
+        migrated,
+        dry_run: false,
+    })
+}
+
 pub async fn run(pool: &PgPool, params: DeleteParams<'_>) -> Result<DeleteResult> {
     match params.name {
-        Some(name) => {
-            let kind = params
-                .kind
-                .ok_or_else(|| anyhow::anyhow!("--kind is required when --name is specified"))?;
-            delete_one(
-                pool,
-                params.namespace,
-                kind,
-                name,
-                params.dry_run,
-                params.user_id,
-            )
-            .await
-        }
+        Some(name) => delete_one(pool, name, params.folder, params.dry_run, params.user_id).await,
         None => {
+            if params.folder.is_none() && params.entry_type.is_none() {
+                anyhow::bail!(
+                    "Bulk delete requires at least one of: name, folder, or type filter."
+                );
+            }
             delete_bulk(
                 pool,
-                params.namespace,
-                params.kind,
+                params.folder,
+                params.entry_type,
                 params.dry_run,
                 params.user_id,
             )
@@ -58,102 +282,190 @@ pub async fn run(pool: &PgPool, params: DeleteParams<'_>) -> Result<DeleteResult
 
 async fn delete_one(
     pool: &PgPool,
-    namespace: &str,
-    kind: &str,
     name: &str,
+    folder: Option<&str>,
     dry_run: bool,
     user_id: Option<Uuid>,
 ) -> Result<DeleteResult> {
     if dry_run {
-        let exists: bool = if let Some(uid) = user_id {
-            sqlx::query_scalar(
-                "SELECT EXISTS(SELECT 1 FROM entries \
-                 WHERE user_id = $1 AND namespace = $2 AND kind = $3 AND name = $4)",
+        // Dry-run uses the same disambiguation logic as actual delete:
+        // - 0 matches  → nothing to delete
+        // - 1 match    → show what would be deleted (with correct folder/type)
+        // - 2+ matches → disambiguation error (same as non-dry-run)
+        #[derive(sqlx::FromRow)]
+        struct DryRunRow {
+            id: Uuid,
+            folder: String,
+            #[sqlx(rename = "type")]
+            entry_type: String,
+        }
+
+        let rows: Vec<DryRunRow> = if let Some(uid) = user_id {
+            if let Some(f) = folder {
+                sqlx::query_as(
+                    "SELECT id, folder, type FROM entries WHERE user_id = $1 AND folder = $2 AND name = $3",
+                )
+                .bind(uid)
+                .bind(f)
+                .bind(name)
+                .fetch_all(pool)
+                .await?
+            } else {
+                sqlx::query_as(
+                    "SELECT id, folder, type FROM entries WHERE user_id = $1 AND name = $2",
+                )
+                .bind(uid)
+                .bind(name)
+                .fetch_all(pool)
+                .await?
+            }
+        } else if let Some(f) = folder {
+            sqlx::query_as(
+                "SELECT id, folder, type FROM entries WHERE user_id IS NULL AND folder = $1 AND name = $2",
             )
-            .bind(uid)
-            .bind(namespace)
-            .bind(kind)
+            .bind(f)
             .bind(name)
-            .fetch_one(pool)
+            .fetch_all(pool)
             .await?
         } else {
-            sqlx::query_scalar(
-                "SELECT EXISTS(SELECT 1 FROM entries \
-                 WHERE user_id IS NULL AND namespace = $1 AND kind = $2 AND name = $3)",
+            sqlx::query_as(
+                "SELECT id, folder, type FROM entries WHERE user_id IS NULL AND name = $1",
             )
-            .bind(namespace)
-            .bind(kind)
             .bind(name)
-            .fetch_one(pool)
+            .fetch_all(pool)
             .await?
         };
 
-        let deleted = if exists {
-            vec![DeletedEntry {
-                namespace: namespace.to_string(),
-                kind: kind.to_string(),
-                name: name.to_string(),
-            }]
-        } else {
-            vec![]
+        return match rows.len() {
+            0 => Ok(DeleteResult {
+                deleted: vec![],
+                migrated: vec![],
+                dry_run: true,
+            }),
+            1 => {
+                let row = rows.into_iter().next().unwrap();
+                let refs =
+                    fetch_key_referrers_pool(pool, row.id, &row.folder, name, user_id).await?;
+                Ok(DeleteResult {
+                    deleted: vec![DeletedEntry {
+                        name: name.to_string(),
+                        folder: row.folder,
+                        entry_type: row.entry_type,
+                    }],
+                    migrated: refs.iter().map(ref_label).collect(),
+                    dry_run: true,
+                })
+            }
+            _ => {
+                let folders: Vec<&str> = rows.iter().map(|r| r.folder.as_str()).collect();
+                anyhow::bail!(
+                    "Ambiguous: {} entries named '{}' found in folders: [{}]. \
+                     Specify 'folder' to disambiguate.",
+                    rows.len(),
+                    name,
+                    folders.join(", ")
+                )
+            }
         };
-        return Ok(DeleteResult {
-            deleted,
-            dry_run: true,
-        });
     }
 
     let mut tx = pool.begin().await?;
 
-    let row: Option<EntryRow> = if let Some(uid) = user_id {
+    // Fetch matching rows with FOR UPDATE; use folder when provided to resolve ambiguity.
+    let rows: Vec<EntryRow> = if let Some(uid) = user_id {
+        if let Some(f) = folder {
+            sqlx::query_as(
+                "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+                 WHERE user_id = $1 AND folder = $2 AND name = $3 FOR UPDATE",
+            )
+            .bind(uid)
+            .bind(f)
+            .bind(name)
+            .fetch_all(&mut *tx)
+            .await?
+        } else {
+            sqlx::query_as(
+                "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+                 WHERE user_id = $1 AND name = $2 FOR UPDATE",
+            )
+            .bind(uid)
+            .bind(name)
+            .fetch_all(&mut *tx)
+            .await?
+        }
+    } else if let Some(f) = folder {
         sqlx::query_as(
-            "SELECT id, version, tags, metadata FROM entries \
-             WHERE user_id = $1 AND namespace = $2 AND kind = $3 AND name = $4 FOR UPDATE",
+            "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+             WHERE user_id IS NULL AND folder = $1 AND name = $2 FOR UPDATE",
         )
-        .bind(uid)
-        .bind(namespace)
-        .bind(kind)
+        .bind(f)
         .bind(name)
-        .fetch_optional(&mut *tx)
+        .fetch_all(&mut *tx)
         .await?
     } else {
         sqlx::query_as(
-            "SELECT id, version, tags, metadata FROM entries \
-             WHERE user_id IS NULL AND namespace = $1 AND kind = $2 AND name = $3 FOR UPDATE",
+            "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+             WHERE user_id IS NULL AND name = $1 FOR UPDATE",
         )
-        .bind(namespace)
-        .bind(kind)
         .bind(name)
-        .fetch_optional(&mut *tx)
+        .fetch_all(&mut *tx)
         .await?
     };
 
-    let Some(row) = row else {
-        tx.rollback().await?;
-        return Ok(DeleteResult {
-            deleted: vec![],
-            dry_run: false,
-        });
+    let row = match rows.len() {
+        0 => {
+            tx.rollback().await?;
+            return Ok(DeleteResult {
+                deleted: vec![],
+                migrated: vec![],
+                dry_run: false,
+            });
+        }
+        1 => rows.into_iter().next().unwrap(),
+        _ => {
+            tx.rollback().await?;
+            let folders: Vec<&str> = rows.iter().map(|r| r.folder.as_str()).collect();
+            anyhow::bail!(
+                "Ambiguous: {} entries named '{}' found in folders: [{}]. \
+                 Specify 'folder' to disambiguate.",
+                rows.len(),
+                name,
+                folders.join(", ")
+            )
+        }
     };
 
-    snapshot_and_delete(&mut tx, namespace, kind, name, &row, user_id).await?;
-    crate::audit::log_tx(&mut tx, user_id, "delete", namespace, kind, name, json!({})).await;
+    let folder = row.folder.clone();
+    let entry_type = row.entry_type.clone();
+    let migrated = migrate_key_refs_if_needed(&mut tx, &row, name, user_id, false).await?;
+    snapshot_and_delete(&mut tx, &folder, &entry_type, name, &row, user_id).await?;
+    crate::audit::log_tx(
+        &mut tx,
+        user_id,
+        "delete",
+        &folder,
+        &entry_type,
+        name,
+        json!({}),
+    )
+    .await;
     tx.commit().await?;
 
     Ok(DeleteResult {
         deleted: vec![DeletedEntry {
-            namespace: namespace.to_string(),
-            kind: kind.to_string(),
             name: name.to_string(),
+            folder,
+            entry_type,
         }],
+        migrated,
         dry_run: false,
     })
 }
 
 async fn delete_bulk(
     pool: &PgPool,
-    namespace: &str,
-    kind: Option<&str>,
+    folder: Option<&str>,
+    entry_type: Option<&str>,
     dry_run: bool,
     user_id: Option<Uuid>,
 ) -> Result<DeleteResult> {
@@ -161,111 +473,125 @@ async fn delete_bulk(
     struct FullEntryRow {
         id: Uuid,
         version: i64,
-        kind: String,
+        folder: String,
+        #[sqlx(rename = "type")]
+        entry_type: String,
         name: String,
         metadata: serde_json::Value,
         tags: Vec<String>,
+        notes: String,
     }
 
-    let rows: Vec<FullEntryRow> = match (user_id, kind) {
-        (Some(uid), Some(k)) => {
-            sqlx::query_as(
-                "SELECT id, version, kind, name, metadata, tags FROM entries \
-             WHERE user_id = $1 AND namespace = $2 AND kind = $3 ORDER BY name",
-            )
-            .bind(uid)
-            .bind(namespace)
-            .bind(k)
-            .fetch_all(pool)
-            .await?
-        }
-        (Some(uid), None) => {
-            sqlx::query_as(
-                "SELECT id, version, kind, name, metadata, tags FROM entries \
-             WHERE user_id = $1 AND namespace = $2 ORDER BY kind, name",
-            )
-            .bind(uid)
-            .bind(namespace)
-            .fetch_all(pool)
-            .await?
-        }
-        (None, Some(k)) => {
-            sqlx::query_as(
-                "SELECT id, version, kind, name, metadata, tags FROM entries \
-             WHERE user_id IS NULL AND namespace = $1 AND kind = $2 ORDER BY name",
-            )
-            .bind(namespace)
-            .bind(k)
-            .fetch_all(pool)
-            .await?
-        }
-        (None, None) => {
-            sqlx::query_as(
-                "SELECT id, version, kind, name, metadata, tags FROM entries \
-             WHERE user_id IS NULL AND namespace = $1 ORDER BY kind, name",
-            )
-            .bind(namespace)
-            .fetch_all(pool)
-            .await?
-        }
-    };
+    let mut conditions: Vec<String> = Vec::new();
+    let mut idx: i32 = 1;
+
+    if user_id.is_some() {
+        conditions.push(format!("user_id = ${}", idx));
+        idx += 1;
+    } else {
+        conditions.push("user_id IS NULL".to_string());
+    }
+    if folder.is_some() {
+        conditions.push(format!("folder = ${}", idx));
+        idx += 1;
+    }
+    if entry_type.is_some() {
+        conditions.push(format!("type = ${}", idx));
+    }
+
+    let where_clause = format!("WHERE {}", conditions.join(" AND "));
+    let sql = format!(
+        "SELECT id, version, folder, type, name, metadata, tags, notes \
+         FROM entries {where_clause} ORDER BY type, name"
+    );
+
+    let mut q = sqlx::query_as::<_, FullEntryRow>(&sql);
+    if let Some(uid) = user_id {
+        q = q.bind(uid);
+    }
+    if let Some(f) = folder {
+        q = q.bind(f);
+    }
+    if let Some(t) = entry_type {
+        q = q.bind(t);
+    }
+    let rows = q.fetch_all(pool).await?;
 
     if dry_run {
+        let mut migrated: Vec<String> = Vec::new();
+        for row in &rows {
+            let refs =
+                fetch_key_referrers_pool(pool, row.id, &row.folder, &row.name, user_id).await?;
+            migrated.extend(refs.iter().map(ref_label));
+        }
         let deleted = rows
             .iter()
             .map(|r| DeletedEntry {
-                namespace: namespace.to_string(),
-                kind: r.kind.clone(),
                 name: r.name.clone(),
+                folder: r.folder.clone(),
+                entry_type: r.entry_type.clone(),
             })
             .collect();
         return Ok(DeleteResult {
             deleted,
+            migrated,
             dry_run: true,
         });
     }
 
     let mut deleted = Vec::with_capacity(rows.len());
+    let mut migrated: Vec<String> = Vec::new();
     for row in &rows {
         let entry_row = EntryRow {
             id: row.id,
             version: row.version,
+            folder: row.folder.clone(),
+            entry_type: row.entry_type.clone(),
             tags: row.tags.clone(),
             metadata: row.metadata.clone(),
+            notes: row.notes.clone(),
         };
         let mut tx = pool.begin().await?;
+        let m = migrate_key_refs_if_needed(&mut tx, &entry_row, &row.name, user_id, false).await?;
+        migrated.extend(m);
         snapshot_and_delete(
-            &mut tx, namespace, &row.kind, &row.name, &entry_row, user_id,
+            &mut tx,
+            &row.folder,
+            &row.entry_type,
+            &row.name,
+            &entry_row,
+            user_id,
         )
         .await?;
         crate::audit::log_tx(
             &mut tx,
             user_id,
             "delete",
-            namespace,
-            &row.kind,
+            &row.folder,
+            &row.entry_type,
             &row.name,
             json!({"bulk": true}),
         )
         .await;
         tx.commit().await?;
         deleted.push(DeletedEntry {
-            namespace: namespace.to_string(),
-            kind: row.kind.clone(),
             name: row.name.clone(),
+            folder: row.folder.clone(),
+            entry_type: row.entry_type.clone(),
         });
     }
 
     Ok(DeleteResult {
         deleted,
+        migrated,
         dry_run: false,
     })
 }
 
 async fn snapshot_and_delete(
     tx: &mut sqlx::Transaction<'_, sqlx::Postgres>,
-    namespace: &str,
-    kind: &str,
+    folder: &str,
+    entry_type: &str,
     name: &str,
     row: &EntryRow,
     user_id: Option<Uuid>,
@@ -275,8 +601,8 @@ async fn snapshot_and_delete(
         db::EntrySnapshotParams {
             entry_id: row.id,
             user_id,
-            namespace,
-            kind,
+            folder,
+            entry_type,
             name,
             version: row.version,
             action: "delete",
@@ -320,3 +646,264 @@ async fn snapshot_and_delete(
 
     Ok(())
 }
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use serde_json::json;
+
+    async fn maybe_test_pool() -> Option<PgPool> {
+        let Ok(url) = std::env::var("SECRETS_DATABASE_URL") else {
+            eprintln!("skip delete migration tests: SECRETS_DATABASE_URL is not set");
+            return None;
+        };
+        let Ok(pool) = PgPool::connect(&url).await else {
+            eprintln!("skip delete migration tests: cannot connect to database");
+            return None;
+        };
+        if let Err(e) = crate::db::migrate(&pool).await {
+            eprintln!("skip delete migration tests: migrate failed: {e}");
+            return None;
+        }
+        Some(pool)
+    }
+
+    async fn insert_entry(
+        pool: &PgPool,
+        id: Uuid,
+        user_id: Uuid,
+        folder: &str,
+        entry_type: &str,
+        name: &str,
+        metadata: serde_json::Value,
+    ) -> Result<()> {
+        sqlx::query(
+            "INSERT INTO entries (id, user_id, folder, type, name, notes, tags, metadata, version) \
+             VALUES ($1, $2, $3, $4, $5, '', ARRAY[]::text[], $6, 1)",
+        )
+        .bind(id)
+        .bind(user_id)
+        .bind(folder)
+        .bind(entry_type)
+        .bind(name)
+        .bind(metadata)
+        .execute(pool)
+        .await?;
+        Ok(())
+    }
+
+    #[tokio::test]
+    async fn delete_shared_key_dry_run_reports_migration_without_writes() -> Result<()> {
+        let Some(pool) = maybe_test_pool().await else {
+            return Ok(());
+        };
+
+        let user_id = Uuid::from_u128(rand::random());
+        let key_id = Uuid::from_u128(rand::random());
+        let ref_a = Uuid::from_u128(rand::random());
+        let ref_b = Uuid::from_u128(rand::random());
+
+        insert_entry(
+            &pool,
+            key_id,
+            user_id,
+            "kfolder",
+            "key",
+            "shared-key",
+            json!({}),
+        )
+        .await?;
+        sqlx::query("INSERT INTO secrets (entry_id, field_name, encrypted) VALUES ($1, $2, $3)")
+            .bind(key_id)
+            .bind("pem")
+            .bind(vec![1_u8, 2, 3])
+            .execute(&pool)
+            .await?;
+
+        insert_entry(
+            &pool,
+            ref_a,
+            user_id,
+            "afolder",
+            "server",
+            "srv-a",
+            json!({"key_ref":"kfolder/shared-key"}),
+        )
+        .await?;
+        insert_entry(
+            &pool,
+            ref_b,
+            user_id,
+            "bfolder",
+            "server",
+            "srv-b",
+            json!({"key_ref":"shared-key"}),
+        )
+        .await?;
+
+        let result = run(
+            &pool,
+            DeleteParams {
+                name: Some("shared-key"),
+                folder: Some("kfolder"),
+                entry_type: None,
+                dry_run: true,
+                user_id: Some(user_id),
+            },
+        )
+        .await?;
+
+        assert!(result.dry_run);
+        assert_eq!(result.deleted.len(), 1);
+        assert_eq!(result.migrated.len(), 2);
+
+        let key_exists: bool = sqlx::query_scalar(
+            "SELECT EXISTS(SELECT 1 FROM entries WHERE id = $1 AND user_id = $2)",
+        )
+        .bind(key_id)
+        .bind(user_id)
+        .fetch_one(&pool)
+        .await?;
+        assert!(key_exists);
+
+        let ref_a_key_ref: Option<String> =
+            sqlx::query_scalar("SELECT metadata->>'key_ref' FROM entries WHERE id = $1")
+                .bind(ref_a)
+                .fetch_one(&pool)
+                .await?;
+        let ref_b_key_ref: Option<String> =
+            sqlx::query_scalar("SELECT metadata->>'key_ref' FROM entries WHERE id = $1")
+                .bind(ref_b)
+                .fetch_one(&pool)
+                .await?;
+        assert_eq!(ref_a_key_ref.as_deref(), Some("kfolder/shared-key"));
+        assert_eq!(ref_b_key_ref.as_deref(), Some("shared-key"));
+
+        sqlx::query("DELETE FROM entries WHERE user_id = $1")
+            .bind(user_id)
+            .execute(&pool)
+            .await?;
+        Ok(())
+    }
+
+    #[tokio::test]
+    async fn delete_shared_key_auto_migrates_single_copy_and_redirects_refs() -> Result<()> {
+        let Some(pool) = maybe_test_pool().await else {
+            return Ok(());
+        };
+
+        let user_id = Uuid::from_u128(rand::random());
+        let key_id = Uuid::from_u128(rand::random());
+        let ref_a = Uuid::from_u128(rand::random());
+        let ref_b = Uuid::from_u128(rand::random());
+        let ref_c = Uuid::from_u128(rand::random());
+
+        insert_entry(
+            &pool,
+            key_id,
+            user_id,
+            "kfolder",
+            "key",
+            "shared-key",
+            json!({}),
+        )
+        .await?;
+        sqlx::query("INSERT INTO secrets (entry_id, field_name, encrypted) VALUES ($1, $2, $3)")
+            .bind(key_id)
+            .bind("pem")
+            .bind(vec![7_u8, 8, 9])
+            .execute(&pool)
+            .await?;
+
+        // owner candidate (sorted first by folder)
+        insert_entry(
+            &pool,
+            ref_a,
+            user_id,
+            "afolder",
+            "server",
+            "srv-a",
+            json!({"key_ref":"kfolder/shared-key"}),
+        )
+        .await?;
+        insert_entry(
+            &pool,
+            ref_b,
+            user_id,
+            "bfolder",
+            "server",
+            "srv-b",
+            json!({"key_ref":"shared-key"}),
+        )
+        .await?;
+        insert_entry(
+            &pool,
+            ref_c,
+            user_id,
+            "cfolder",
+            "service",
+            "svc-c",
+            json!({"key_ref":"kfolder/shared-key"}),
+        )
+        .await?;
+
+        let result = run(
+            &pool,
+            DeleteParams {
+                name: Some("shared-key"),
+                folder: Some("kfolder"),
+                entry_type: None,
+                dry_run: false,
+                user_id: Some(user_id),
+            },
+        )
+        .await?;
+
+        assert!(!result.dry_run);
+        assert_eq!(result.deleted.len(), 1);
+        assert_eq!(result.migrated.len(), 3);
+
+        let key_exists: bool = sqlx::query_scalar(
+            "SELECT EXISTS(SELECT 1 FROM entries WHERE id = $1 AND user_id = $2)",
+        )
+        .bind(key_id)
+        .bind(user_id)
+        .fetch_one(&pool)
+        .await?;
+        assert!(!key_exists);
+
+        let owner_key_ref: Option<String> =
+            sqlx::query_scalar("SELECT metadata->>'key_ref' FROM entries WHERE id = $1")
+                .bind(ref_a)
+                .fetch_one(&pool)
+                .await?;
+        let ref_b_key_ref: Option<String> =
+            sqlx::query_scalar("SELECT metadata->>'key_ref' FROM entries WHERE id = $1")
+                .bind(ref_b)
+                .fetch_one(&pool)
+                .await?;
+        let ref_c_key_ref: Option<String> =
+            sqlx::query_scalar("SELECT metadata->>'key_ref' FROM entries WHERE id = $1")
+                .bind(ref_c)
+                .fetch_one(&pool)
+                .await?;
+
+        assert_eq!(owner_key_ref, None);
+        assert_eq!(ref_b_key_ref.as_deref(), Some("afolder/srv-a"));
+        assert_eq!(ref_c_key_ref.as_deref(), Some("afolder/srv-a"));
+
+        let owner_has_copied: bool = sqlx::query_scalar(
+            "SELECT EXISTS(SELECT 1 FROM secrets WHERE entry_id = $1 AND field_name = 'pem')",
+        )
+        .bind(ref_a)
+        .fetch_one(&pool)
+        .await?;
+        assert!(owner_has_copied);
+
+        sqlx::query("DELETE FROM entries WHERE user_id = $1")
+            .bind(user_id)
+            .execute(&pool)
+            .await?;
+        Ok(())
+    }
+}

crates/secrets-core/src/service/env_map.rs

@@ -12,8 +12,8 @@ use crate::service::search::{fetch_entries, fetch_secrets_for_entries};
 #[allow(clippy::too_many_arguments)]
 pub async fn build_env_map(
     pool: &PgPool,
-    namespace: Option<&str>,
-    kind: Option<&str>,
+    folder: Option<&str>,
+    entry_type: Option<&str>,
     name: Option<&str>,
     tags: &[String],
     only_fields: &[String],
@@ -21,12 +21,13 @@ pub async fn build_env_map(
     master_key: &[u8; 32],
     user_id: Option<Uuid>,
 ) -> Result<HashMap<String, String>> {
-    let entries = fetch_entries(pool, namespace, kind, name, tags, None, user_id).await?;
+    let entries = fetch_entries(pool, folder, entry_type, name, tags, None, user_id).await?;
 
     let mut combined: HashMap<String, String> = HashMap::new();
 
     for entry in &entries {
-        let entry_map = build_entry_env_map(pool, entry, only_fields, prefix, master_key).await?;
+        let entry_map =
+            build_entry_env_map(pool, entry, only_fields, prefix, master_key, user_id).await?;
         combined.extend(entry_map);
     }
 
@@ -39,6 +40,7 @@ async fn build_entry_env_map(
     only_fields: &[String],
     prefix: &str,
     master_key: &[u8; 32],
+    user_id: Option<Uuid>,
 ) -> Result<HashMap<String, String>> {
     let entry_ids = vec![entry.id];
     let secrets_map = fetch_secrets_for_entries(pool, &entry_ids).await?;
@@ -66,18 +68,23 @@ async fn build_entry_env_map(
         map.insert(key, json_to_env_string(&decrypted));
     }
 
-    // Resolve key_ref
+    // Resolve key_ref. Supported formats: "name" or "folder/name".
     if let Some(key_ref) = entry.metadata.get("key_ref").and_then(|v| v.as_str()) {
-        let key_entries = fetch_entries(
-            pool,
-            Some(&entry.namespace),
-            Some("key"),
-            Some(key_ref),
-            &[],
-            None,
-            None,
-        )
-        .await?;
+        let (ref_folder, ref_name) = if let Some((f, n)) = key_ref.split_once('/') {
+            (Some(f), n)
+        } else {
+            (None, key_ref)
+        };
+        let key_entries =
+            fetch_entries(pool, ref_folder, None, Some(ref_name), &[], None, user_id).await?;
+
+        if key_entries.len() > 1 {
+            anyhow::bail!(
+                "key_ref '{}' matched {} entries; qualify with folder/name to resolve the ambiguity",
+                key_ref,
+                key_entries.len()
+            );
+        }
 
         if let Some(key_entry) = key_entries.first() {
             let key_ids = vec![key_entry.id];
@@ -95,7 +102,7 @@ async fn build_entry_env_map(
                 map.insert(key_var, json_to_env_string(&decrypted));
             }
         } else {
-            tracing::warn!(key_ref, "key_ref target not found");
+            tracing::warn!(key_ref, ?user_id, "key_ref target not found");
         }
     }
 

crates/secrets-core/src/service/export.rs

@@ -9,8 +9,8 @@ use crate::models::{ExportData, ExportEntry, ExportFormat};
 use crate::service::search::{fetch_entries, fetch_secrets_for_entries};
 
 pub struct ExportParams<'a> {
-    pub namespace: Option<&'a str>,
-    pub kind: Option<&'a str>,
+    pub folder: Option<&'a str>,
+    pub entry_type: Option<&'a str>,
     pub name: Option<&'a str>,
     pub tags: &'a [String],
     pub query: Option<&'a str>,
@@ -25,8 +25,8 @@ pub async fn export(
 ) -> Result<ExportData> {
     let entries = fetch_entries(
         pool,
-        params.namespace,
-        params.kind,
+        params.folder,
+        params.entry_type,
         params.name,
         params.tags,
         params.query,
@@ -62,9 +62,10 @@ pub async fn export(
         };
 
         export_entries.push(ExportEntry {
-            namespace: entry.namespace.clone(),
-            kind: entry.kind.clone(),
             name: entry.name.clone(),
+            folder: entry.folder.clone(),
+            entry_type: entry.entry_type.clone(),
+            notes: entry.notes.clone(),
             tags: entry.tags.clone(),
             metadata: entry.metadata.clone(),
             secrets,

crates/secrets-core/src/service/get_secret.rs

@@ -5,31 +5,19 @@ use std::collections::HashMap;
 use uuid::Uuid;
 
 use crate::crypto;
-use crate::service::search::{fetch_entries, fetch_secrets_for_entries};
+use crate::service::search::{fetch_secrets_for_entries, resolve_entry, resolve_entry_by_id};
 
 /// Decrypt a single named field from an entry.
+/// `folder` is optional; if omitted and multiple entries share the name, an error is returned.
 pub async fn get_secret_field(
     pool: &PgPool,
-    namespace: &str,
-    kind: &str,
     name: &str,
+    folder: Option<&str>,
     field_name: &str,
     master_key: &[u8; 32],
     user_id: Option<Uuid>,
 ) -> Result<Value> {
-    let entries = fetch_entries(
-        pool,
-        Some(namespace),
-        Some(kind),
-        Some(name),
-        &[],
-        None,
-        user_id,
-    )
-    .await?;
-    let entry = entries
-        .first()
-        .ok_or_else(|| anyhow::anyhow!("Not found: [{}/{}] {}", namespace, kind, name))?;
+    let entry = resolve_entry(pool, name, folder, user_id).await?;
 
     let entry_ids = vec![entry.id];
     let secrets_map = fetch_secrets_for_entries(pool, &entry_ids).await?;
@@ -44,27 +32,15 @@ pub async fn get_secret_field(
 }
 
 /// Decrypt all secret fields from an entry. Returns a map field_name → decrypted Value.
+/// `folder` is optional; if omitted and multiple entries share the name, an error is returned.
 pub async fn get_all_secrets(
     pool: &PgPool,
-    namespace: &str,
-    kind: &str,
     name: &str,
+    folder: Option<&str>,
     master_key: &[u8; 32],
     user_id: Option<Uuid>,
 ) -> Result<HashMap<String, Value>> {
-    let entries = fetch_entries(
-        pool,
-        Some(namespace),
-        Some(kind),
-        Some(name),
-        &[],
-        None,
-        user_id,
-    )
-    .await?;
-    let entry = entries
-        .first()
-        .ok_or_else(|| anyhow::anyhow!("Not found: [{}/{}] {}", namespace, kind, name))?;
+    let entry = resolve_entry(pool, name, folder, user_id).await?;
 
     let entry_ids = vec![entry.id];
     let secrets_map = fetch_secrets_for_entries(pool, &entry_ids).await?;
@@ -77,3 +53,52 @@ pub async fn get_all_secrets(
     }
     Ok(map)
 }
+
+/// Decrypt a single named field from an entry, located by its UUID.
+pub async fn get_secret_field_by_id(
+    pool: &PgPool,
+    entry_id: Uuid,
+    field_name: &str,
+    master_key: &[u8; 32],
+    user_id: Option<Uuid>,
+) -> Result<Value> {
+    resolve_entry_by_id(pool, entry_id, user_id)
+        .await
+        .map_err(|_| anyhow::anyhow!("Entry with id '{}' not found", entry_id))?;
+
+    let entry_ids = vec![entry_id];
+    let secrets_map = fetch_secrets_for_entries(pool, &entry_ids).await?;
+    let fields = secrets_map.get(&entry_id).map(Vec::as_slice).unwrap_or(&[]);
+
+    let field = fields
+        .iter()
+        .find(|f| f.field_name == field_name)
+        .ok_or_else(|| anyhow::anyhow!("Secret field '{}' not found", field_name))?;
+
+    crypto::decrypt_json(master_key, &field.encrypted)
+}
+
+/// Decrypt all secret fields from an entry, located by its UUID.
+/// Returns a map field_name → decrypted Value.
+pub async fn get_all_secrets_by_id(
+    pool: &PgPool,
+    entry_id: Uuid,
+    master_key: &[u8; 32],
+    user_id: Option<Uuid>,
+) -> Result<HashMap<String, Value>> {
+    // Validate entry exists (and that it belongs to the requesting user)
+    resolve_entry_by_id(pool, entry_id, user_id)
+        .await
+        .map_err(|_| anyhow::anyhow!("Entry with id '{}' not found", entry_id))?;
+
+    let entry_ids = vec![entry_id];
+    let secrets_map = fetch_secrets_for_entries(pool, &entry_ids).await?;
+    let fields = secrets_map.get(&entry_id).map(Vec::as_slice).unwrap_or(&[]);
+
+    let mut map = HashMap::new();
+    for f in fields {
+        let decrypted = crypto::decrypt_json(master_key, &f.encrypted)?;
+        map.insert(f.field_name.clone(), decrypted);
+    }
+    Ok(map)
+}

crates/secrets-core/src/service/history.rs

@@ -3,6 +3,8 @@ use serde_json::Value;
 use sqlx::PgPool;
 use uuid::Uuid;
 
+use crate::service::search::resolve_entry;
+
 #[derive(Debug, serde::Serialize)]
 pub struct HistoryEntry {
     pub version: i64,
@@ -10,11 +12,12 @@ pub struct HistoryEntry {
     pub created_at: String,
 }
 
+/// Return version history for the entry identified by `name`.
+/// `folder` is optional; if omitted and multiple entries share the name, an error is returned.
 pub async fn run(
     pool: &PgPool,
-    namespace: &str,
-    kind: &str,
     name: &str,
+    folder: Option<&str>,
     limit: u32,
     user_id: Option<Uuid>,
 ) -> Result<Vec<HistoryEntry>> {
@@ -25,32 +28,16 @@ pub async fn run(
         created_at: chrono::DateTime<chrono::Utc>,
     }
 
-    let rows: Vec<Row> = if let Some(uid) = user_id {
-        sqlx::query_as(
-            "SELECT version, action, created_at FROM entries_history \
-             WHERE namespace = $1 AND kind = $2 AND name = $3 AND user_id = $4 \
-             ORDER BY id DESC LIMIT $5",
-        )
-        .bind(namespace)
-        .bind(kind)
-        .bind(name)
-        .bind(uid)
-        .bind(limit as i64)
-        .fetch_all(pool)
-        .await?
-    } else {
-        sqlx::query_as(
-            "SELECT version, action, created_at FROM entries_history \
-             WHERE namespace = $1 AND kind = $2 AND name = $3 AND user_id IS NULL \
-             ORDER BY id DESC LIMIT $4",
-        )
-        .bind(namespace)
-        .bind(kind)
-        .bind(name)
-        .bind(limit as i64)
-        .fetch_all(pool)
-        .await?
-    };
+    let entry = resolve_entry(pool, name, folder, user_id).await?;
+
+    let rows: Vec<Row> = sqlx::query_as(
+        "SELECT version, action, created_at FROM entries_history \
+         WHERE entry_id = $1 ORDER BY id DESC LIMIT $2",
+    )
+    .bind(entry.id)
+    .bind(limit as i64)
+    .fetch_all(pool)
+    .await?;
 
     Ok(rows
         .into_iter()
@@ -64,12 +51,11 @@ pub async fn run(
 
 pub async fn run_json(
     pool: &PgPool,
-    namespace: &str,
-    kind: &str,
     name: &str,
+    folder: Option<&str>,
     limit: u32,
     user_id: Option<Uuid>,
 ) -> Result<Value> {
-    let entries = run(pool, namespace, kind, name, limit, user_id).await?;
+    let entries = run(pool, name, folder, limit, user_id).await?;
     Ok(serde_json::to_value(entries)?)
 }

crates/secrets-core/src/service/import.rs

@@ -47,10 +47,9 @@ pub async fn run(
     for entry in &data.entries {
         let exists: bool = sqlx::query_scalar(
             "SELECT EXISTS(SELECT 1 FROM entries \
-             WHERE namespace = $1 AND kind = $2 AND name = $3 AND user_id IS NOT DISTINCT FROM $4)",
+             WHERE folder = $1 AND name = $2 AND user_id IS NOT DISTINCT FROM $3)",
         )
-        .bind(&entry.namespace)
-        .bind(&entry.kind)
+        .bind(&entry.folder)
         .bind(&entry.name)
         .bind(params.user_id)
         .fetch_one(pool)
@@ -59,9 +58,7 @@ pub async fn run(
 
         if exists && !params.force {
             return Err(anyhow::anyhow!(
-                "Import aborted: conflict on [{}/{}/{}]",
-                entry.namespace,
-                entry.kind,
+                "Import aborted: conflict on '{}'",
                 entry.name
             ));
         }
@@ -81,9 +78,10 @@ pub async fn run(
         match add_run(
             pool,
             AddParams {
-                namespace: &entry.namespace,
-                kind: &entry.kind,
                 name: &entry.name,
+                folder: &entry.folder,
+                entry_type: &entry.entry_type,
+                notes: &entry.notes,
                 tags: &entry.tags,
                 meta_entries: &meta_entries,
                 secret_entries: &secret_entries,
@@ -98,8 +96,6 @@ pub async fn run(
             }
             Err(e) => {
                 tracing::error!(
-                    namespace = entry.namespace,
-                    kind = entry.kind,
                     name = entry.name,
                     error = %e,
                     "failed to import entry"

crates/secrets-core/src/service/rollback.rs

@@ -8,17 +8,19 @@ use crate::db;
 
 #[derive(Debug, serde::Serialize)]
 pub struct RollbackResult {
-    pub namespace: String,
-    pub kind: String,
     pub name: String,
+    pub folder: String,
+    #[serde(rename = "type")]
+    pub entry_type: String,
     pub restored_version: i64,
 }
 
+/// Roll back entry `name` to `to_version` (or the most recent snapshot if None).
+/// `folder` is optional; if omitted and multiple entries share the name, an error is returned.
 pub async fn run(
     pool: &PgPool,
-    namespace: &str,
-    kind: &str,
     name: &str,
+    folder: Option<&str>,
     to_version: Option<i64>,
     master_key: &[u8; 32],
     user_id: Option<Uuid>,
@@ -26,69 +28,122 @@ pub async fn run(
     #[derive(sqlx::FromRow)]
     struct EntryHistoryRow {
         entry_id: Uuid,
+        folder: String,
+        #[sqlx(rename = "type")]
+        entry_type: String,
         version: i64,
         action: String,
         tags: Vec<String>,
         metadata: Value,
     }
 
-    let snap: Option<EntryHistoryRow> = if let Some(ver) = to_version {
-        if let Some(uid) = user_id {
-            sqlx::query_as(
-                "SELECT entry_id, version, action, tags, metadata FROM entries_history \
-                 WHERE namespace = $1 AND kind = $2 AND name = $3 AND version = $4 \
-                 AND user_id = $5 ORDER BY id DESC LIMIT 1",
+    // Disambiguate: find the unique entry_id for (name, folder).
+    // Query entries_history by entry_id once we know it; first resolve via name + optional folder.
+    let entry_id: Option<Uuid> = if let Some(uid) = user_id {
+        if let Some(f) = folder {
+            sqlx::query_scalar(
+                "SELECT DISTINCT entry_id FROM entries_history \
+                 WHERE name = $1 AND folder = $2 AND user_id = $3 LIMIT 1",
             )
-            .bind(namespace)
-            .bind(kind)
             .bind(name)
-            .bind(ver)
+            .bind(f)
             .bind(uid)
             .fetch_optional(pool)
             .await?
         } else {
-            sqlx::query_as(
-                "SELECT entry_id, version, action, tags, metadata FROM entries_history \
-                 WHERE namespace = $1 AND kind = $2 AND name = $3 AND version = $4 \
-                 AND user_id IS NULL ORDER BY id DESC LIMIT 1",
+            let ids: Vec<Uuid> = sqlx::query_scalar(
+                "SELECT DISTINCT entry_id FROM entries_history \
+                 WHERE name = $1 AND user_id = $2",
             )
-            .bind(namespace)
-            .bind(kind)
             .bind(name)
-            .bind(ver)
-            .fetch_optional(pool)
-            .await?
+            .bind(uid)
+            .fetch_all(pool)
+            .await?;
+            match ids.len() {
+                0 => None,
+                1 => Some(ids[0]),
+                _ => {
+                    let folders: Vec<String> = sqlx::query_scalar(
+                        "SELECT DISTINCT folder FROM entries_history \
+                         WHERE name = $1 AND user_id = $2",
+                    )
+                    .bind(name)
+                    .bind(uid)
+                    .fetch_all(pool)
+                    .await?;
+                    anyhow::bail!(
+                        "Ambiguous: entries named '{}' exist in folders: [{}]. \
+                         Specify 'folder' to disambiguate.",
+                        name,
+                        folders.join(", ")
+                    )
+                }
+            }
         }
-    } else if let Some(uid) = user_id {
-        sqlx::query_as(
-            "SELECT entry_id, version, action, tags, metadata FROM entries_history \
-             WHERE namespace = $1 AND kind = $2 AND name = $3 \
-             AND user_id = $4 ORDER BY id DESC LIMIT 1",
+    } else if let Some(f) = folder {
+        sqlx::query_scalar(
+            "SELECT DISTINCT entry_id FROM entries_history \
+             WHERE name = $1 AND folder = $2 AND user_id IS NULL LIMIT 1",
         )
-        .bind(namespace)
-        .bind(kind)
         .bind(name)
-        .bind(uid)
+        .bind(f)
+        .fetch_optional(pool)
+        .await?
+    } else {
+        let ids: Vec<Uuid> = sqlx::query_scalar(
+            "SELECT DISTINCT entry_id FROM entries_history \
+             WHERE name = $1 AND user_id IS NULL",
+        )
+        .bind(name)
+        .fetch_all(pool)
+        .await?;
+        match ids.len() {
+            0 => None,
+            1 => Some(ids[0]),
+            _ => {
+                let folders: Vec<String> = sqlx::query_scalar(
+                    "SELECT DISTINCT folder FROM entries_history \
+                     WHERE name = $1 AND user_id IS NULL",
+                )
+                .bind(name)
+                .fetch_all(pool)
+                .await?;
+                anyhow::bail!(
+                    "Ambiguous: entries named '{}' exist in folders: [{}]. \
+                     Specify 'folder' to disambiguate.",
+                    name,
+                    folders.join(", ")
+                )
+            }
+        }
+    };
+
+    let entry_id = entry_id.ok_or_else(|| anyhow::anyhow!("No history found for '{}'", name))?;
+
+    let snap: Option<EntryHistoryRow> = if let Some(ver) = to_version {
+        sqlx::query_as(
+            "SELECT entry_id, folder, type, version, action, tags, metadata \
+             FROM entries_history \
+             WHERE entry_id = $1 AND version = $2 ORDER BY id DESC LIMIT 1",
+        )
+        .bind(entry_id)
+        .bind(ver)
         .fetch_optional(pool)
         .await?
     } else {
         sqlx::query_as(
-            "SELECT entry_id, version, action, tags, metadata FROM entries_history \
-             WHERE namespace = $1 AND kind = $2 AND name = $3 \
-             AND user_id IS NULL ORDER BY id DESC LIMIT 1",
+            "SELECT entry_id, folder, type, version, action, tags, metadata \
+             FROM entries_history \
+             WHERE entry_id = $1 ORDER BY id DESC LIMIT 1",
         )
-        .bind(namespace)
-        .bind(kind)
-        .bind(name)
+        .bind(entry_id)
         .fetch_optional(pool)
         .await?
     };
 
     let snap = snap.ok_or_else(|| {
         anyhow::anyhow!(
-            "No history found for [{}/{}] {}{}.",
-            namespace,
-            kind,
+            "No history found for '{}'{}.",
             name,
             to_version
                 .map(|v| format!(" at version {}", v))
@@ -130,43 +185,32 @@ pub async fn run(
     struct LiveEntry {
         id: Uuid,
         version: i64,
+        folder: String,
+        #[sqlx(rename = "type")]
+        entry_type: String,
         tags: Vec<String>,
         metadata: Value,
+        #[allow(dead_code)]
+        notes: String,
     }
 
-    // Query live entry with correct user_id scoping to avoid PK conflicts
-    let live: Option<LiveEntry> = if let Some(uid) = user_id {
-        sqlx::query_as(
-            "SELECT id, version, tags, metadata FROM entries \
-             WHERE user_id = $1 AND namespace = $2 AND kind = $3 AND name = $4 FOR UPDATE",
-        )
-        .bind(uid)
-        .bind(namespace)
-        .bind(kind)
-        .bind(name)
-        .fetch_optional(&mut *tx)
-        .await?
-    } else {
-        sqlx::query_as(
-            "SELECT id, version, tags, metadata FROM entries \
-             WHERE user_id IS NULL AND namespace = $1 AND kind = $2 AND name = $3 FOR UPDATE",
-        )
-        .bind(namespace)
-        .bind(kind)
-        .bind(name)
-        .fetch_optional(&mut *tx)
-        .await?
-    };
+    // Lock the live entry if it exists (matched by entry_id for precision).
+    let live: Option<LiveEntry> = sqlx::query_as(
+        "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+         WHERE id = $1 FOR UPDATE",
+    )
+    .bind(entry_id)
+    .fetch_optional(&mut *tx)
+    .await?;
 
-    let entry_id = if let Some(ref lr) = live {
-        // Snapshot current state before overwriting
+    let live_entry_id = if let Some(ref lr) = live {
         if let Err(e) = db::snapshot_entry_history(
             &mut tx,
             db::EntrySnapshotParams {
                 entry_id: lr.id,
                 user_id,
-                namespace,
-                kind,
+                folder: &lr.folder,
+                entry_type: &lr.entry_type,
                 name,
                 version: lr.version,
                 action: "rollback",
@@ -209,7 +253,6 @@ pub async fn run(
             }
         }
 
-        // Update the existing row in-place to preserve its primary key and user_id
         sqlx::query(
             "UPDATE entries SET tags = $1, metadata = $2, version = version + 1, \
              updated_at = NOW() WHERE id = $3",
@@ -222,16 +265,15 @@ pub async fn run(
 
         lr.id
     } else {
-        // No live entry — insert a fresh one with a new UUID
         if let Some(uid) = user_id {
             sqlx::query_scalar(
                 "INSERT INTO entries \
-                 (user_id, namespace, kind, name, tags, metadata, version, updated_at) \
-                 VALUES ($1, $2, $3, $4, $5, $6, $7, NOW()) RETURNING id",
+                 (user_id, folder, type, name, notes, tags, metadata, version, updated_at) \
+                 VALUES ($1, $2, $3, $4, '', $5, $6, $7, NOW()) RETURNING id",
             )
             .bind(uid)
-            .bind(namespace)
-            .bind(kind)
+            .bind(&snap.folder)
+            .bind(&snap.entry_type)
             .bind(name)
             .bind(&snap.tags)
             .bind(&snap.metadata)
@@ -241,11 +283,11 @@ pub async fn run(
         } else {
             sqlx::query_scalar(
                 "INSERT INTO entries \
-                 (namespace, kind, name, tags, metadata, version, updated_at) \
-                 VALUES ($1, $2, $3, $4, $5, $6, NOW()) RETURNING id",
+                 (folder, type, name, notes, tags, metadata, version, updated_at) \
+                 VALUES ($1, $2, $3, '', $4, $5, $6, NOW()) RETURNING id",
             )
-            .bind(namespace)
-            .bind(kind)
+            .bind(&snap.folder)
+            .bind(&snap.entry_type)
             .bind(name)
             .bind(&snap.tags)
             .bind(&snap.metadata)
@@ -256,7 +298,7 @@ pub async fn run(
     };
 
     sqlx::query("DELETE FROM secrets WHERE entry_id = $1")
-        .bind(entry_id)
+        .bind(live_entry_id)
         .execute(&mut *tx)
         .await?;
 
@@ -265,7 +307,7 @@ pub async fn run(
             continue;
         }
         sqlx::query("INSERT INTO secrets (entry_id, field_name, encrypted) VALUES ($1, $2, $3)")
-            .bind(entry_id)
+            .bind(live_entry_id)
             .bind(&f.field_name)
             .bind(&f.encrypted)
             .execute(&mut *tx)
@@ -276,8 +318,8 @@ pub async fn run(
         &mut tx,
         user_id,
         "rollback",
-        namespace,
-        kind,
+        &snap.folder,
+        &snap.entry_type,
         name,
         serde_json::json!({
             "restored_version": snap.version,
@@ -289,9 +331,9 @@ pub async fn run(
     tx.commit().await?;
 
     Ok(RollbackResult {
-        namespace: namespace.to_string(),
-        kind: kind.to_string(),
         name: name.to_string(),
+        folder: snap.folder,
+        entry_type: snap.entry_type,
         restored_version: snap.version,
     })
 }

crates/secrets-core/src/service/search.rs

@@ -9,8 +9,8 @@ use crate::models::{Entry, SecretField};
 pub const FETCH_ALL_LIMIT: u32 = 100_000;
 
 pub struct SearchParams<'a> {
-    pub namespace: Option<&'a str>,
-    pub kind: Option<&'a str>,
+    pub folder: Option<&'a str>,
+    pub entry_type: Option<&'a str>,
     pub name: Option<&'a str>,
     pub tags: &'a [String],
     pub query: Option<&'a str>,
@@ -27,49 +27,46 @@ pub struct SearchResult {
     pub secret_schemas: HashMap<Uuid, Vec<SecretField>>,
 }
 
-pub async fn run(pool: &PgPool, params: SearchParams<'_>) -> Result<SearchResult> {
-    let entries = fetch_entries_paged(pool, &params).await?;
-    let entry_ids: Vec<Uuid> = entries.iter().map(|e| e.id).collect();
-    let secret_schemas = if !entry_ids.is_empty() {
-        fetch_secret_schemas(pool, &entry_ids).await?
-    } else {
-        HashMap::new()
-    };
-    Ok(SearchResult {
-        entries,
-        secret_schemas,
-    })
-}
-
-/// Fetch entries matching the given filters — returns all matching entries up to FETCH_ALL_LIMIT.
-pub async fn fetch_entries(
-    pool: &PgPool,
-    namespace: Option<&str>,
-    kind: Option<&str>,
-    name: Option<&str>,
-    tags: &[String],
-    query: Option<&str>,
-    user_id: Option<Uuid>,
-) -> Result<Vec<Entry>> {
-    let params = SearchParams {
-        namespace,
-        kind,
-        name,
-        tags,
-        query,
-        sort: "name",
-        limit: FETCH_ALL_LIMIT,
-        offset: 0,
-        user_id,
-    };
+/// List `entries` rows matching params (paged, ordered per `params.sort`).
+/// Does not read the `secrets` table.
+pub async fn list_entries(pool: &PgPool, params: SearchParams<'_>) -> Result<Vec<Entry>> {
     fetch_entries_paged(pool, &params).await
 }
 
-async fn fetch_entries_paged(pool: &PgPool, a: &SearchParams<'_>) -> Result<Vec<Entry>> {
+/// Count `entries` rows matching the same filters as [`list_entries`] (ignores `sort` / `limit` / `offset`).
+/// Does not read the `secrets` table.
+pub async fn count_entries(pool: &PgPool, a: &SearchParams<'_>) -> Result<i64> {
+    let (where_clause, _) = entry_where_clause_and_next_idx(a);
+    let sql = format!("SELECT COUNT(*)::bigint FROM entries {where_clause}");
+    let mut q = sqlx::query_scalar::<_, i64>(&sql);
+    if let Some(uid) = a.user_id {
+        q = q.bind(uid);
+    }
+    if let Some(v) = a.folder {
+        q = q.bind(v);
+    }
+    if let Some(v) = a.entry_type {
+        q = q.bind(v);
+    }
+    if let Some(v) = a.name {
+        q = q.bind(v);
+    }
+    for tag in a.tags {
+        q = q.bind(tag);
+    }
+    if let Some(v) = a.query {
+        let pattern = format!("%{}%", v.replace('%', "\\%").replace('_', "\\_"));
+        q = q.bind(pattern);
+    }
+    let n = q.fetch_one(pool).await?;
+    Ok(n)
+}
+
+/// Shared WHERE clause and the next `$n` index (for LIMIT/OFFSET in paged queries).
+fn entry_where_clause_and_next_idx(a: &SearchParams<'_>) -> (String, i32) {
     let mut conditions: Vec<String> = Vec::new();
     let mut idx: i32 = 1;
 
-    // user_id filtering — always comes first when present
     if a.user_id.is_some() {
         conditions.push(format!("user_id = ${}", idx));
         idx += 1;
@@ -77,12 +74,12 @@ async fn fetch_entries_paged(pool: &PgPool, a: &SearchParams<'_>) -> Result<Vec<
         conditions.push("user_id IS NULL".to_string());
     }
 
-    if a.namespace.is_some() {
-        conditions.push(format!("namespace = ${}", idx));
+    if a.folder.is_some() {
+        conditions.push(format!("folder = ${}", idx));
         idx += 1;
     }
-    if a.kind.is_some() {
-        conditions.push(format!("kind = ${}", idx));
+    if a.entry_type.is_some() {
+        conditions.push(format!("type = ${}", idx));
         idx += 1;
     }
     if a.name.is_some() {
@@ -106,14 +103,64 @@ async fn fetch_entries_paged(pool: &PgPool, a: &SearchParams<'_>) -> Result<Vec<
     }
     if a.query.is_some() {
         conditions.push(format!(
-            "(name ILIKE ${i} ESCAPE '\\' OR namespace ILIKE ${i} ESCAPE '\\' \
-             OR kind ILIKE ${i} ESCAPE '\\' OR metadata::text ILIKE ${i} ESCAPE '\\' \
+            "(name ILIKE ${i} ESCAPE '\\' OR folder ILIKE ${i} ESCAPE '\\' \
+             OR type ILIKE ${i} ESCAPE '\\' OR notes ILIKE ${i} ESCAPE '\\' \
+             OR metadata::text ILIKE ${i} ESCAPE '\\' \
              OR EXISTS (SELECT 1 FROM unnest(tags) t WHERE t ILIKE ${i} ESCAPE '\\'))",
             i = idx
         ));
         idx += 1;
     }
 
+    let where_clause = if conditions.is_empty() {
+        String::new()
+    } else {
+        format!("WHERE {}", conditions.join(" AND "))
+    };
+    (where_clause, idx)
+}
+
+pub async fn run(pool: &PgPool, params: SearchParams<'_>) -> Result<SearchResult> {
+    let entries = fetch_entries_paged(pool, &params).await?;
+    let entry_ids: Vec<Uuid> = entries.iter().map(|e| e.id).collect();
+    let secret_schemas = if !entry_ids.is_empty() {
+        fetch_secret_schemas(pool, &entry_ids).await?
+    } else {
+        HashMap::new()
+    };
+    Ok(SearchResult {
+        entries,
+        secret_schemas,
+    })
+}
+
+/// Fetch entries matching the given filters — returns all matching entries up to FETCH_ALL_LIMIT.
+pub async fn fetch_entries(
+    pool: &PgPool,
+    folder: Option<&str>,
+    entry_type: Option<&str>,
+    name: Option<&str>,
+    tags: &[String],
+    query: Option<&str>,
+    user_id: Option<Uuid>,
+) -> Result<Vec<Entry>> {
+    let params = SearchParams {
+        folder,
+        entry_type,
+        name,
+        tags,
+        query,
+        sort: "name",
+        limit: FETCH_ALL_LIMIT,
+        offset: 0,
+        user_id,
+    };
+    list_entries(pool, params).await
+}
+
+async fn fetch_entries_paged(pool: &PgPool, a: &SearchParams<'_>) -> Result<Vec<Entry>> {
+    let (where_clause, idx) = entry_where_clause_and_next_idx(a);
+
     let order = match a.sort {
         "updated" => "updated_at DESC",
         "created" => "created_at DESC",
@@ -121,30 +168,22 @@ async fn fetch_entries_paged(pool: &PgPool, a: &SearchParams<'_>) -> Result<Vec<
     };
 
     let limit_idx = idx;
-    idx += 1;
-    let offset_idx = idx;
-
-    let where_clause = if conditions.is_empty() {
-        String::new()
-    } else {
-        format!("WHERE {}", conditions.join(" AND "))
-    };
+    let offset_idx = idx + 1;
 
     let sql = format!(
-        "SELECT id, user_id, \
-         namespace, kind, name, tags, metadata, version, created_at, updated_at \
+        "SELECT id, user_id, folder, type, name, notes, tags, metadata, version, \
+         created_at, updated_at \
          FROM entries {where_clause} ORDER BY {order} LIMIT ${limit_idx} OFFSET ${offset_idx}"
     );
 
     let mut q = sqlx::query_as::<_, EntryRaw>(&sql);
-
     if let Some(uid) = a.user_id {
         q = q.bind(uid);
     }
-    if let Some(v) = a.namespace {
+    if let Some(v) = a.folder {
         q = q.bind(v);
     }
-    if let Some(v) = a.kind {
+    if let Some(v) = a.entry_type {
         q = q.bind(v);
     }
     if let Some(v) = a.name {
@@ -207,15 +246,81 @@ pub async fn fetch_secrets_for_entries(
     Ok(map)
 }
 
-// ── Internal raw row (because user_id is nullable in DB) ─────────────────────
+/// Resolve exactly one entry by its UUID primary key.
+///
+/// Returns an error if the entry does not exist or does not belong to the given user.
+pub async fn resolve_entry_by_id(
+    pool: &PgPool,
+    id: Uuid,
+    user_id: Option<Uuid>,
+) -> Result<crate::models::Entry> {
+    let row: Option<EntryRaw> = if let Some(uid) = user_id {
+        sqlx::query_as(
+            "SELECT id, user_id, folder, type, name, notes, tags, metadata, version, \
+             created_at, updated_at FROM entries WHERE id = $1 AND user_id = $2",
+        )
+        .bind(id)
+        .bind(uid)
+        .fetch_optional(pool)
+        .await?
+    } else {
+        sqlx::query_as(
+            "SELECT id, user_id, folder, type, name, notes, tags, metadata, version, \
+             created_at, updated_at FROM entries WHERE id = $1 AND user_id IS NULL",
+        )
+        .bind(id)
+        .fetch_optional(pool)
+        .await?
+    };
+    row.map(Entry::from)
+        .ok_or_else(|| anyhow::anyhow!("Entry with id '{}' not found", id))
+}
 
+/// Resolve exactly one entry by name, with optional folder for disambiguation.
+///
+/// - If `folder` is provided: exact `(folder, name)` match.
+/// - If `folder` is None and exactly one entry matches: returns it.
+/// - If `folder` is None and multiple entries match: returns an error listing
+///   the folders and asking the caller to specify one.
+pub async fn resolve_entry(
+    pool: &PgPool,
+    name: &str,
+    folder: Option<&str>,
+    user_id: Option<Uuid>,
+) -> Result<crate::models::Entry> {
+    let entries = fetch_entries(pool, folder, None, Some(name), &[], None, user_id).await?;
+    match entries.len() {
+        0 => {
+            if let Some(f) = folder {
+                anyhow::bail!("Not found: '{}' in folder '{}'", name, f)
+            } else {
+                anyhow::bail!("Not found: '{}'", name)
+            }
+        }
+        1 => Ok(entries.into_iter().next().unwrap()),
+        _ => {
+            let folders: Vec<&str> = entries.iter().map(|e| e.folder.as_str()).collect();
+            anyhow::bail!(
+                "Ambiguous: {} entries named '{}' found in folders: [{}]. \
+                 Specify 'folder' to disambiguate.",
+                entries.len(),
+                name,
+                folders.join(", ")
+            )
+        }
+    }
+}
+
+// ── Internal raw row (because user_id is nullable in DB) ─────────────────────
 #[derive(sqlx::FromRow)]
 struct EntryRaw {
     id: Uuid,
     user_id: Option<Uuid>,
-    namespace: String,
-    kind: String,
+    folder: String,
+    #[sqlx(rename = "type")]
+    entry_type: String,
     name: String,
+    notes: String,
     tags: Vec<String>,
     metadata: Value,
     version: i64,
@@ -228,9 +333,10 @@ impl From<EntryRaw> for Entry {
         Entry {
             id: r.id,
             user_id: r.user_id,
-            namespace: r.namespace,
-            kind: r.kind,
+            folder: r.folder,
+            entry_type: r.entry_type,
             name: r.name,
+            notes: r.notes,
             tags: r.tags,
             metadata: r.metadata,
             version: r.version,

crates/secrets-core/src/service/update.rs

@@ -5,7 +5,7 @@ use uuid::Uuid;
 
 use crate::crypto;
 use crate::db;
-use crate::models::EntryRow;
+use crate::models::{EntryRow, EntryWriteRow};
 use crate::service::add::{
     collect_field_paths, collect_key_paths, flatten_json_fields, insert_path, parse_key_path,
     parse_kv, remove_path,
@@ -13,9 +13,10 @@ use crate::service::add::{
 
 #[derive(Debug, serde::Serialize)]
 pub struct UpdateResult {
-    pub namespace: String,
-    pub kind: String,
     pub name: String,
+    pub folder: String,
+    #[serde(rename = "type")]
+    pub entry_type: String,
     pub add_tags: Vec<String>,
     pub remove_tags: Vec<String>,
     pub meta_keys: Vec<String>,
@@ -25,9 +26,10 @@ pub struct UpdateResult {
 }
 
 pub struct UpdateParams<'a> {
-    pub namespace: &'a str,
-    pub kind: &'a str,
     pub name: &'a str,
+    /// Optional folder for disambiguation when multiple entries share the same name.
+    pub folder: Option<&'a str>,
+    pub notes: Option<&'a str>,
     pub add_tags: &'a [String],
     pub remove_tags: &'a [String],
     pub meta_entries: &'a [String],
@@ -44,45 +46,76 @@ pub async fn run(
 ) -> Result<UpdateResult> {
     let mut tx = pool.begin().await?;
 
-    let row: Option<EntryRow> = if let Some(uid) = params.user_id {
+    // Fetch matching rows with FOR UPDATE; use folder when provided to resolve ambiguity.
+    let rows: Vec<EntryRow> = if let Some(uid) = params.user_id {
+        if let Some(folder) = params.folder {
+            sqlx::query_as(
+                "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+                 WHERE user_id = $1 AND folder = $2 AND name = $3 FOR UPDATE",
+            )
+            .bind(uid)
+            .bind(folder)
+            .bind(params.name)
+            .fetch_all(&mut *tx)
+            .await?
+        } else {
+            sqlx::query_as(
+                "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+                 WHERE user_id = $1 AND name = $2 FOR UPDATE",
+            )
+            .bind(uid)
+            .bind(params.name)
+            .fetch_all(&mut *tx)
+            .await?
+        }
+    } else if let Some(folder) = params.folder {
         sqlx::query_as(
-            "SELECT id, version, tags, metadata FROM entries \
-             WHERE user_id = $1 AND namespace = $2 AND kind = $3 AND name = $4 FOR UPDATE",
+            "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+             WHERE user_id IS NULL AND folder = $1 AND name = $2 FOR UPDATE",
         )
-        .bind(uid)
-        .bind(params.namespace)
-        .bind(params.kind)
+        .bind(folder)
         .bind(params.name)
-        .fetch_optional(&mut *tx)
+        .fetch_all(&mut *tx)
         .await?
     } else {
         sqlx::query_as(
-            "SELECT id, version, tags, metadata FROM entries \
-             WHERE user_id IS NULL AND namespace = $1 AND kind = $2 AND name = $3 FOR UPDATE",
+            "SELECT id, version, folder, type, tags, metadata, notes FROM entries \
+             WHERE user_id IS NULL AND name = $1 FOR UPDATE",
         )
-        .bind(params.namespace)
-        .bind(params.kind)
         .bind(params.name)
-        .fetch_optional(&mut *tx)
+        .fetch_all(&mut *tx)
         .await?
     };
 
-    let row = row.ok_or_else(|| {
-        anyhow::anyhow!(
-            "Not found: [{}/{}] {}. Use `add` to create it first.",
-            params.namespace,
-            params.kind,
-            params.name
-        )
-    })?;
+    let row = match rows.len() {
+        0 => {
+            tx.rollback().await?;
+            anyhow::bail!(
+                "Not found: '{}'. Use `add` to create it first.",
+                params.name
+            )
+        }
+        1 => rows.into_iter().next().unwrap(),
+        _ => {
+            tx.rollback().await?;
+            let folders: Vec<&str> = rows.iter().map(|r| r.folder.as_str()).collect();
+            anyhow::bail!(
+                "Ambiguous: {} entries named '{}' found in folders: [{}]. \
+                 Specify 'folder' to disambiguate.",
+                rows.len(),
+                params.name,
+                folders.join(", ")
+            )
+        }
+    };
 
     if let Err(e) = db::snapshot_entry_history(
         &mut tx,
         db::EntrySnapshotParams {
             entry_id: row.id,
             user_id: params.user_id,
-            namespace: params.namespace,
-            kind: params.kind,
+            folder: &row.folder,
+            entry_type: &row.entry_type,
             name: params.name,
             version: row.version,
             action: "update",
@@ -117,12 +150,16 @@ pub async fn run(
     }
     let metadata = Value::Object(meta_map);
 
+    let new_notes = params.notes.unwrap_or(&row.notes);
+
     let result = sqlx::query(
-        "UPDATE entries SET tags = $1, metadata = $2, version = version + 1, updated_at = NOW() \
-         WHERE id = $3 AND version = $4",
+        "UPDATE entries SET tags = $1, metadata = $2, notes = $3, \
+         version = version + 1, updated_at = NOW() \
+         WHERE id = $4 AND version = $5",
     )
     .bind(&tags)
     .bind(&metadata)
+    .bind(new_notes)
     .bind(row.id)
     .bind(row.version)
     .execute(&mut *tx)
@@ -131,9 +168,7 @@ pub async fn run(
     if result.rows_affected() == 0 {
         tx.rollback().await?;
         anyhow::bail!(
-            "Concurrent modification detected for [{}/{}] {}. Please retry.",
-            params.namespace,
-            params.kind,
+            "Concurrent modification detected for '{}'. Please retry.",
             params.name
         );
     }
@@ -243,8 +278,8 @@ pub async fn run(
         &mut tx,
         params.user_id,
         "update",
-        params.namespace,
-        params.kind,
+        "",
+        "",
         params.name,
         serde_json::json!({
             "add_tags": params.add_tags,
@@ -260,9 +295,9 @@ pub async fn run(
     tx.commit().await?;
 
     Ok(UpdateResult {
-        namespace: params.namespace.to_string(),
-        kind: params.kind.to_string(),
         name: params.name.to_string(),
+        folder: row.folder.clone(),
+        entry_type: row.entry_type.clone(),
         add_tags: params.add_tags.to_vec(),
         remove_tags: params.remove_tags.to_vec(),
         meta_keys,
@@ -271,3 +306,118 @@ pub async fn run(
         remove_secrets: remove_secret_keys,
     })
 }
+
+/// Update non-sensitive entry columns by primary key (multi-tenant: `user_id` must match).
+/// Does not read or modify `secrets` rows.
+pub struct UpdateEntryFieldsByIdParams<'a> {
+    pub folder: &'a str,
+    pub entry_type: &'a str,
+    pub name: &'a str,
+    pub notes: &'a str,
+    pub tags: &'a [String],
+    pub metadata: &'a serde_json::Value,
+}
+
+pub async fn update_fields_by_id(
+    pool: &PgPool,
+    entry_id: Uuid,
+    user_id: Uuid,
+    params: UpdateEntryFieldsByIdParams<'_>,
+) -> Result<()> {
+    if params.folder.len() > 128 {
+        anyhow::bail!("folder must be at most 128 characters");
+    }
+    if params.entry_type.len() > 64 {
+        anyhow::bail!("type must be at most 64 characters");
+    }
+    if params.name.len() > 256 {
+        anyhow::bail!("name must be at most 256 characters");
+    }
+
+    let mut tx = pool.begin().await?;
+
+    let row: Option<EntryWriteRow> = sqlx::query_as(
+        "SELECT id, version, folder, type, name, tags, metadata, notes FROM entries \
+         WHERE id = $1 AND user_id = $2 FOR UPDATE",
+    )
+    .bind(entry_id)
+    .bind(user_id)
+    .fetch_optional(&mut *tx)
+    .await?;
+
+    let row = match row {
+        Some(r) => r,
+        None => {
+            tx.rollback().await?;
+            anyhow::bail!("Entry not found");
+        }
+    };
+
+    if let Err(e) = db::snapshot_entry_history(
+        &mut tx,
+        db::EntrySnapshotParams {
+            entry_id: row.id,
+            user_id: Some(user_id),
+            folder: &row.folder,
+            entry_type: &row.entry_type,
+            name: &row.name,
+            version: row.version,
+            action: "update",
+            tags: &row.tags,
+            metadata: &row.metadata,
+        },
+    )
+    .await
+    {
+        tracing::warn!(error = %e, "failed to snapshot entry history before web update");
+    }
+
+    let res = sqlx::query(
+        "UPDATE entries SET folder = $1, type = $2, name = $3, notes = $4, tags = $5, metadata = $6, \
+         version = version + 1, updated_at = NOW() \
+         WHERE id = $7 AND version = $8",
+    )
+    .bind(params.folder)
+    .bind(params.entry_type)
+    .bind(params.name)
+    .bind(params.notes)
+    .bind(params.tags)
+    .bind(params.metadata)
+    .bind(row.id)
+    .bind(row.version)
+    .execute(&mut *tx)
+    .await
+    .map_err(|e| {
+        if let sqlx::Error::Database(ref d) = e
+            && d.code().as_deref() == Some("23505")
+        {
+            return anyhow::anyhow!(
+                "An entry with this folder and name already exists for your account."
+            );
+        }
+        e.into()
+    })?;
+
+    if res.rows_affected() == 0 {
+        tx.rollback().await?;
+        anyhow::bail!("Concurrent modification detected. Please refresh and try again.");
+    }
+
+    crate::audit::log_tx(
+        &mut tx,
+        Some(user_id),
+        "update",
+        params.folder,
+        params.entry_type,
+        params.name,
+        serde_json::json!({
+            "source": "web",
+            "entry_id": entry_id,
+            "fields": ["folder", "type", "name", "notes", "tags", "metadata"],
+        }),
+    )
+    .await;
+
+    tx.commit().await?;
+    Ok(())
+}

crates/secrets-mcp/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "secrets-mcp"
-version = "0.2.2"
+version = "0.3.7"
 edition.workspace = true
 
 [[bin]]

crates/secrets-mcp/src/main.rs

@@ -21,7 +21,7 @@ use tower_sessions_sqlx_store_chrono::PostgresStore;
 use tracing_subscriber::EnvFilter;
 use tracing_subscriber::fmt::time::FormatTime;
 
-use secrets_core::config::resolve_db_url;
+use secrets_core::config::resolve_db_config;
 use secrets_core::db::{create_pool, migrate};
 
 use crate::oauth::OAuthConfig;
@@ -78,9 +78,9 @@ async fn main() -> Result<()> {
         .init();
 
     // ── Database ──────────────────────────────────────────────────────────────
-    let db_url = resolve_db_url("")
+    let db_config = resolve_db_config("")
         .context("Database not configured. Set SECRETS_DATABASE_URL environment variable.")?;
-    let pool = create_pool(&db_url)
+    let pool = create_pool(&db_config)
         .await
         .context("failed to connect to database")?;
     migrate(&pool)

crates/secrets-mcp/src/tools.rs

@@ -14,6 +14,7 @@ use rmcp::{
 };
 use schemars::JsonSchema;
 use serde::Deserialize;
+use serde_json::{Map, Value};
 use sqlx::PgPool;
 use uuid::Uuid;
 
@@ -22,10 +23,10 @@ use secrets_core::service::{
     add::{AddParams, run as svc_add},
     delete::{DeleteParams, run as svc_delete},
     export::{ExportParams, export as svc_export},
-    get_secret::{get_all_secrets, get_secret_field},
+    get_secret::{get_all_secrets_by_id, get_secret_field_by_id},
     history::run as svc_history,
     rollback::run as svc_rollback,
-    search::{SearchParams, run as svc_search},
+    search::{SearchParams, resolve_entry_by_id, run as svc_search},
     update::{UpdateParams, run as svc_update},
 };
 
@@ -154,18 +155,38 @@ impl SecretsService {
 // ── Tool parameter types ──────────────────────────────────────────────────────
 
 #[derive(Debug, Deserialize, JsonSchema)]
-struct SearchInput {
-    #[schemars(description = "Namespace filter (e.g. 'refining', 'ricnsmart')")]
-    namespace: Option<String>,
-    #[schemars(description = "Kind filter (e.g. 'server', 'service', 'key')")]
-    kind: Option<String>,
-    #[schemars(description = "Exact record name")]
+struct FindInput {
+    #[schemars(
+        description = "Fuzzy search across name, folder, type, notes, tags, and metadata values"
+    )]
+    query: Option<String>,
+    #[schemars(description = "Exact folder filter (e.g. 'refining', 'ricnsmart')")]
+    folder: Option<String>,
+    #[schemars(description = "Exact type filter (e.g. 'server', 'service', 'person', 'key')")]
+    #[serde(rename = "type")]
+    entry_type: Option<String>,
+    #[schemars(description = "Exact name filter")]
     name: Option<String>,
     #[schemars(description = "Tag filters (all must match)")]
     tags: Option<Vec<String>>,
-    #[schemars(description = "Fuzzy search across name, namespace, kind, tags, metadata")]
+    #[schemars(description = "Max results (default 20)")]
+    limit: Option<u32>,
+}
+
+#[derive(Debug, Deserialize, JsonSchema)]
+struct SearchInput {
+    #[schemars(description = "Fuzzy search across name, folder, type, notes, tags, metadata")]
     query: Option<String>,
-    #[schemars(description = "Return only summary fields (name/tags/desc/updated_at)")]
+    #[schemars(description = "Folder filter (e.g. 'refining', 'personal', 'family')")]
+    folder: Option<String>,
+    #[schemars(description = "Type filter (e.g. 'server', 'service', 'person', 'key')")]
+    #[serde(rename = "type")]
+    entry_type: Option<String>,
+    #[schemars(description = "Exact name to match")]
+    name: Option<String>,
+    #[schemars(description = "Tag filters (all must match)")]
+    tags: Option<Vec<String>>,
+    #[schemars(description = "Return only summary fields (name/tags/notes/updated_at)")]
     summary: Option<bool>,
     #[schemars(description = "Sort order: 'name' (default), 'updated', 'created'")]
     sort: Option<String>,
@@ -177,96 +198,135 @@ struct SearchInput {
 
 #[derive(Debug, Deserialize, JsonSchema)]
 struct GetSecretInput {
-    #[schemars(description = "Namespace of the entry")]
-    namespace: String,
-    #[schemars(description = "Kind of the entry (e.g. 'server', 'service')")]
-    kind: String,
-    #[schemars(description = "Name of the entry")]
-    name: String,
+    #[schemars(description = "Entry UUID obtained from secrets_find results")]
+    id: String,
     #[schemars(description = "Specific field to retrieve. If omitted, returns all fields.")]
     field: Option<String>,
 }
 
 #[derive(Debug, Deserialize, JsonSchema)]
 struct AddInput {
-    #[schemars(description = "Namespace")]
-    namespace: String,
-    #[schemars(description = "Kind (e.g. 'server', 'service', 'key')")]
-    kind: String,
-    #[schemars(description = "Unique name within namespace+kind")]
+    #[schemars(description = "Unique name for this entry")]
     name: String,
+    #[schemars(description = "Folder for organization (optional, e.g. 'personal', 'refining')")]
+    folder: Option<String>,
+    #[schemars(
+        description = "Type/category of this entry (optional, e.g. 'server', 'person', 'key')"
+    )]
+    #[serde(rename = "type")]
+    entry_type: Option<String>,
+    #[schemars(description = "Free-text notes for this entry (optional)")]
+    notes: Option<String>,
     #[schemars(description = "Tags for this entry")]
     tags: Option<Vec<String>>,
     #[schemars(description = "Metadata fields as 'key=value' or 'key:=json' strings")]
     meta: Option<Vec<String>>,
+    #[schemars(
+        description = "Metadata fields as a JSON object {\"key\": value}. Merged with 'meta' if both provided."
+    )]
+    meta_obj: Option<Map<String, Value>>,
     #[schemars(description = "Secret fields as 'key=value' strings")]
     secrets: Option<Vec<String>>,
+    #[schemars(
+        description = "Secret fields as a JSON object {\"key\": \"value\"}. Merged with 'secrets' if both provided."
+    )]
+    secrets_obj: Option<Map<String, Value>>,
 }
 
 #[derive(Debug, Deserialize, JsonSchema)]
 struct UpdateInput {
-    #[schemars(description = "Namespace")]
-    namespace: String,
-    #[schemars(description = "Kind")]
-    kind: String,
-    #[schemars(description = "Name")]
+    #[schemars(description = "Name of the entry to update")]
     name: String,
+    #[schemars(
+        description = "Folder for disambiguation when multiple entries share the same name (optional)"
+    )]
+    folder: Option<String>,
+    #[schemars(
+        description = "Entry UUID (from secrets_find). If provided, name/folder are used for disambiguation only."
+    )]
+    id: Option<String>,
+    #[schemars(description = "Update the notes field")]
+    notes: Option<String>,
     #[schemars(description = "Tags to add")]
     add_tags: Option<Vec<String>>,
     #[schemars(description = "Tags to remove")]
     remove_tags: Option<Vec<String>>,
     #[schemars(description = "Metadata fields to update/add as 'key=value' strings")]
     meta: Option<Vec<String>>,
+    #[schemars(
+        description = "Metadata fields to update/add as a JSON object {\"key\": value}. Merged with 'meta' if both provided."
+    )]
+    meta_obj: Option<Map<String, Value>>,
     #[schemars(description = "Metadata field keys to remove")]
     remove_meta: Option<Vec<String>>,
     #[schemars(description = "Secret fields to update/add as 'key=value' strings")]
     secrets: Option<Vec<String>>,
+    #[schemars(
+        description = "Secret fields to update/add as a JSON object {\"key\": \"value\"}. Merged with 'secrets' if both provided."
+    )]
+    secrets_obj: Option<Map<String, Value>>,
     #[schemars(description = "Secret field keys to remove")]
     remove_secrets: Option<Vec<String>>,
 }
 
 #[derive(Debug, Deserialize, JsonSchema)]
 struct DeleteInput {
-    #[schemars(description = "Namespace")]
-    namespace: String,
-    #[schemars(description = "Kind filter (required for single delete)")]
-    kind: Option<String>,
-    #[schemars(description = "Exact name to delete. Omit for bulk delete by namespace+kind.")]
+    #[schemars(
+        description = "Entry UUID (from secrets_find). If provided, deletes this specific entry \
+        regardless of name/folder."
+    )]
+    id: Option<String>,
+    #[schemars(description = "Name of the entry to delete (single delete). \
+        Omit to bulk delete by folder/type filters.")]
     name: Option<String>,
+    #[schemars(description = "Folder filter for bulk delete")]
+    folder: Option<String>,
+    #[schemars(description = "Type filter for bulk delete")]
+    #[serde(rename = "type")]
+    entry_type: Option<String>,
     #[schemars(description = "Preview deletions without writing")]
     dry_run: Option<bool>,
 }
 
 #[derive(Debug, Deserialize, JsonSchema)]
 struct HistoryInput {
-    #[schemars(description = "Namespace")]
-    namespace: String,
-    #[schemars(description = "Kind")]
-    kind: String,
-    #[schemars(description = "Name")]
+    #[schemars(description = "Name of the entry")]
     name: String,
+    #[schemars(
+        description = "Folder for disambiguation when multiple entries share the same name (optional)"
+    )]
+    folder: Option<String>,
+    #[schemars(
+        description = "Entry UUID (from secrets_find). If provided, name/folder are ignored."
+    )]
+    id: Option<String>,
     #[schemars(description = "Max history entries to return (default 20)")]
     limit: Option<u32>,
 }
 
 #[derive(Debug, Deserialize, JsonSchema)]
 struct RollbackInput {
-    #[schemars(description = "Namespace")]
-    namespace: String,
-    #[schemars(description = "Kind")]
-    kind: String,
-    #[schemars(description = "Name")]
+    #[schemars(description = "Name of the entry")]
     name: String,
+    #[schemars(
+        description = "Folder for disambiguation when multiple entries share the same name (optional)"
+    )]
+    folder: Option<String>,
+    #[schemars(
+        description = "Entry UUID (from secrets_find). If provided, name/folder are ignored."
+    )]
+    id: Option<String>,
     #[schemars(description = "Target version number. Omit to restore the most recent snapshot.")]
     to_version: Option<i64>,
 }
 
 #[derive(Debug, Deserialize, JsonSchema)]
 struct ExportInput {
-    #[schemars(description = "Namespace filter")]
-    namespace: Option<String>,
-    #[schemars(description = "Kind filter")]
-    kind: Option<String>,
+    #[schemars(description = "Folder filter")]
+    folder: Option<String>,
+    #[schemars(description = "Type filter")]
+    #[serde(rename = "type")]
+    entry_type: Option<String>,
     #[schemars(description = "Exact name filter")]
     name: Option<String>,
     #[schemars(description = "Tag filters")]
@@ -279,27 +339,129 @@ struct ExportInput {
 
 #[derive(Debug, Deserialize, JsonSchema)]
 struct EnvMapInput {
-    #[schemars(description = "Namespace filter")]
-    namespace: Option<String>,
-    #[schemars(description = "Kind filter")]
-    kind: Option<String>,
+    #[schemars(description = "Folder filter")]
+    folder: Option<String>,
+    #[schemars(description = "Type filter")]
+    #[serde(rename = "type")]
+    entry_type: Option<String>,
     #[schemars(description = "Exact name filter")]
     name: Option<String>,
     #[schemars(description = "Tag filters")]
     tags: Option<Vec<String>>,
     #[schemars(description = "Only include these secret fields")]
     only_fields: Option<Vec<String>>,
-    #[schemars(description = "Environment variable name prefix")]
+    #[schemars(description = "Environment variable name prefix. \
+        Variable names are built as UPPER(prefix)_UPPER(entry_name)_UPPER(field_name), \
+        with hyphens and dots replaced by underscores. \
+        Example: entry 'aliyun', field 'access_key_id' → ALIYUN_ACCESS_KEY_ID \
+        (or PREFIX_ALIYUN_ACCESS_KEY_ID with prefix set).")]
     prefix: Option<String>,
 }
 
+#[derive(Debug, Deserialize, JsonSchema)]
+struct OverviewInput {}
+
+// ── Helpers ───────────────────────────────────────────────────────────────────
+
+/// Convert a JSON object map into "key=value" / "key:=json" strings for service-layer parsing.
+fn map_to_kv_strings(map: Map<String, Value>) -> Vec<String> {
+    map.into_iter()
+        .map(|(k, v)| match &v {
+            Value::String(s) => format!("{}={}", k, s),
+            _ => format!("{}:={}", k, v),
+        })
+        .collect()
+}
+
+/// Parse a UUID string, returning an MCP error on failure.
+fn parse_uuid(s: &str) -> Result<Uuid, rmcp::ErrorData> {
+    s.parse::<Uuid>()
+        .map_err(|_| rmcp::ErrorData::invalid_request(format!("Invalid UUID: '{}'", s), None))
+}
+
 // ── Tool implementations ──────────────────────────────────────────────────────
 
 #[tool_router]
 impl SecretsService {
+    #[tool(
+        description = "Find entries in the secrets store by folder, name, type, tags, or a \
+        fuzzy query that also searches metadata values. Requires Bearer API key. \
+        Returns 0 or more entries with id, metadata, and secret field names (not values). \
+        Use the returned id with secrets_get to decrypt secret values. \
+        Replaces secrets_search for discovery tasks.",
+        annotations(title = "Find Secrets", read_only_hint = true, idempotent_hint = true)
+    )]
+    async fn secrets_find(
+        &self,
+        Parameters(input): Parameters<FindInput>,
+        ctx: RequestContext<RoleServer>,
+    ) -> Result<CallToolResult, rmcp::ErrorData> {
+        let t = Instant::now();
+        let user_id = Self::require_user_id(&ctx)?;
+        tracing::info!(
+            tool = "secrets_find",
+            ?user_id,
+            folder = input.folder.as_deref(),
+            entry_type = input.entry_type.as_deref(),
+            name = input.name.as_deref(),
+            query = input.query.as_deref(),
+            "tool call start",
+        );
+        let tags = input.tags.unwrap_or_default();
+        let result = svc_search(
+            &self.pool,
+            SearchParams {
+                folder: input.folder.as_deref(),
+                entry_type: input.entry_type.as_deref(),
+                name: input.name.as_deref(),
+                tags: &tags,
+                query: input.query.as_deref(),
+                sort: "name",
+                limit: input.limit.unwrap_or(20),
+                offset: 0,
+                user_id: Some(user_id),
+            },
+        )
+        .await
+        .map_err(|e| mcp_err_internal_logged("secrets_find", Some(user_id), e))?;
+
+        let entries: Vec<serde_json::Value> = result
+            .entries
+            .iter()
+            .map(|e| {
+                let schema: Vec<&str> = result
+                    .secret_schemas
+                    .get(&e.id)
+                    .map(|f| f.iter().map(|s| s.field_name.as_str()).collect())
+                    .unwrap_or_default();
+                serde_json::json!({
+                    "id": e.id,
+                    "name": e.name,
+                    "folder": e.folder,
+                    "type": e.entry_type,
+                    "tags": e.tags,
+                    "metadata": e.metadata,
+                    "secret_fields": schema,
+                    "updated_at": e.updated_at.format("%Y-%m-%dT%H:%M:%SZ").to_string(),
+                })
+            })
+            .collect();
+
+        tracing::info!(
+            tool = "secrets_find",
+            ?user_id,
+            result_count = entries.len(),
+            elapsed_ms = t.elapsed().as_millis(),
+            "tool call ok",
+        );
+        let json = serde_json::to_string_pretty(&entries).unwrap_or_else(|_| "[]".to_string());
+        Ok(CallToolResult::success(vec![Content::text(json)]))
+    }
+
     #[tool(
         description = "Search entries in the secrets store. Requires Bearer API key. Returns \
-        entries with metadata and secret field names (not values). Use secrets_get to decrypt secret values.",
+        entries with metadata and secret field names (not values). \
+        Prefer secrets_find for discovery; secrets_search is kept for backward compatibility.",
         annotations(
             title = "Search Secrets",
             read_only_hint = true,
@@ -316,8 +478,8 @@ impl SecretsService {
         tracing::info!(
             tool = "secrets_search",
             ?user_id,
-            namespace = input.namespace.as_deref(),
-            kind = input.kind.as_deref(),
+            folder = input.folder.as_deref(),
+            entry_type = input.entry_type.as_deref(),
             name = input.name.as_deref(),
             query = input.query.as_deref(),
             "tool call start",
@@ -326,8 +488,8 @@ impl SecretsService {
         let result = svc_search(
             &self.pool,
             SearchParams {
-                namespace: input.namespace.as_deref(),
-                kind: input.kind.as_deref(),
+                folder: input.folder.as_deref(),
+                entry_type: input.entry_type.as_deref(),
                 name: input.name.as_deref(),
                 tags: &tags,
                 query: input.query.as_deref(),
@@ -347,12 +509,11 @@ impl SecretsService {
             .map(|e| {
                 if summary {
                     serde_json::json!({
-                        "namespace": e.namespace,
-                        "kind": e.kind,
                         "name": e.name,
+                        "folder": e.folder,
+                        "type": e.entry_type,
                         "tags": e.tags,
-                        "desc": e.metadata.get("desc").or_else(|| e.metadata.get("url"))
-                            .and_then(|v| v.as_str()).unwrap_or(""),
+                        "notes": e.notes,
                         "updated_at": e.updated_at.format("%Y-%m-%dT%H:%M:%SZ").to_string(),
                     })
                 } else {
@@ -363,9 +524,10 @@ impl SecretsService {
                         .unwrap_or_default();
                     serde_json::json!({
                         "id": e.id,
-                        "namespace": e.namespace,
-                        "kind": e.kind,
                         "name": e.name,
+                        "folder": e.folder,
+                        "type": e.entry_type,
+                        "notes": e.notes,
                         "tags": e.tags,
                         "metadata": e.metadata,
                         "secret_fields": schema,
@@ -389,8 +551,8 @@ impl SecretsService {
     }
 
     #[tool(
-        description = "Get decrypted secret field values for an entry. Requires your \
-        encryption key via X-Encryption-Key header (64 hex chars, PBKDF2-derived). \
+        description = "Get decrypted secret field values for an entry identified by its UUID \
+        (from secrets_find). Requires X-Encryption-Key header. \
         Returns all fields, or a specific field if 'field' is provided.",
         annotations(
             title = "Get Secret Values",
@@ -405,32 +567,23 @@ impl SecretsService {
     ) -> Result<CallToolResult, rmcp::ErrorData> {
         let t = Instant::now();
         let (user_id, user_key) = Self::require_user_and_key(&ctx)?;
+        let entry_id = parse_uuid(&input.id)?;
         tracing::info!(
             tool = "secrets_get",
-            ?user_id,
-            namespace = %input.namespace,
-            kind = %input.kind,
-            name = %input.name,
+            id = %input.id,
             field = input.field.as_deref(),
             "tool call start",
         );
 
         if let Some(field_name) = &input.field {
-            let value = get_secret_field(
-                &self.pool,
-                &input.namespace,
-                &input.kind,
-                &input.name,
-                field_name,
-                &user_key,
-                Some(user_id),
-            )
-            .await
-            .map_err(|e| mcp_err_internal_logged("secrets_get", Some(user_id), e))?;
+            let value =
+                get_secret_field_by_id(&self.pool, entry_id, field_name, &user_key, Some(user_id))
+                    .await
+                    .map_err(|e| mcp_err_internal_logged("secrets_get", None, e))?;
 
             tracing::info!(
                 tool = "secrets_get",
-                ?user_id,
+                id = %input.id,
                 elapsed_ms = t.elapsed().as_millis(),
                 "tool call ok",
             );
@@ -438,22 +591,14 @@ impl SecretsService {
             let json = serde_json::to_string_pretty(&result).unwrap_or_default();
             Ok(CallToolResult::success(vec![Content::text(json)]))
         } else {
-            let secrets = get_all_secrets(
-                &self.pool,
-                &input.namespace,
-                &input.kind,
-                &input.name,
-                &user_key,
-                Some(user_id),
-            )
-            .await
-            .map_err(|e| mcp_err_internal_logged("secrets_get", Some(user_id), e))?;
+            let secrets = get_all_secrets_by_id(&self.pool, entry_id, &user_key, Some(user_id))
+                .await
+                .map_err(|e| mcp_err_internal_logged("secrets_get", None, e))?;
 
-            let count = secrets.len();
             tracing::info!(
-                tool = "secrets_get",
-                ?user_id,
-                field_count = count,
+            tool = "secrets_get",
+            id = %entry_id,
+            field_count = secrets.len(),
                 elapsed_ms = t.elapsed().as_millis(),
                 "tool call ok",
             );
@@ -465,7 +610,8 @@ impl SecretsService {
     #[tool(
         description = "Add or upsert an entry with metadata and encrypted secret fields. \
         Requires X-Encryption-Key header. \
-        Meta and secret values use 'key=value', 'key=@file', or 'key:=<json>' format.",
+        Meta and secret values use 'key=value', 'key=@file', or 'key:=<json>' format, \
+        or pass a JSON object via meta_obj / secrets_obj.",
         annotations(title = "Add Secret Entry")
     )]
     async fn secrets_add(
@@ -478,22 +624,32 @@ impl SecretsService {
         tracing::info!(
             tool = "secrets_add",
             ?user_id,
-            namespace = %input.namespace,
-            kind = %input.kind,
             name = %input.name,
+            folder = input.folder.as_deref(),
+            entry_type = input.entry_type.as_deref(),
             "tool call start",
         );
 
         let tags = input.tags.unwrap_or_default();
-        let meta = input.meta.unwrap_or_default();
-        let secrets = input.secrets.unwrap_or_default();
+        let mut meta = input.meta.unwrap_or_default();
+        if let Some(obj) = input.meta_obj {
+            meta.extend(map_to_kv_strings(obj));
+        }
+        let mut secrets = input.secrets.unwrap_or_default();
+        if let Some(obj) = input.secrets_obj {
+            secrets.extend(map_to_kv_strings(obj));
+        }
+        let folder = input.folder.as_deref().unwrap_or("");
+        let entry_type = input.entry_type.as_deref().unwrap_or("");
+        let notes = input.notes.as_deref().unwrap_or("");
 
         let result = svc_add(
             &self.pool,
             AddParams {
-                namespace: &input.namespace,
-                kind: &input.kind,
                 name: &input.name,
+                folder,
+                entry_type,
+                notes,
                 tags: &tags,
                 meta_entries: &meta,
                 secret_entries: &secrets,
@@ -507,8 +663,6 @@ impl SecretsService {
         tracing::info!(
             tool = "secrets_add",
             ?user_id,
-            namespace = %input.namespace,
-            kind = %input.kind,
             name = %input.name,
             elapsed_ms = t.elapsed().as_millis(),
             "tool call ok",
@@ -519,7 +673,8 @@ impl SecretsService {
 
     #[tool(
         description = "Incrementally update an existing entry. Requires X-Encryption-Key header. \
-        Only the fields you specify are changed; everything else is preserved.",
+        Only the fields you specify are changed; everything else is preserved. \
+        Optionally pass 'id' (from secrets_find) to target the entry directly.",
         annotations(title = "Update Secret Entry")
     )]
     async fn secrets_update(
@@ -532,25 +687,42 @@ impl SecretsService {
         tracing::info!(
             tool = "secrets_update",
             ?user_id,
-            namespace = %input.namespace,
-            kind = %input.kind,
             name = %input.name,
+            id = ?input.id,
             "tool call start",
         );
 
+        // When id is provided, resolve to (name, folder) via primary key to skip disambiguation.
+        let (resolved_name, resolved_folder): (String, Option<String>) =
+            if let Some(ref id_str) = input.id {
+                let eid = parse_uuid(id_str)?;
+                let entry = resolve_entry_by_id(&self.pool, eid, Some(user_id))
+                    .await
+                    .map_err(|e| mcp_err_internal_logged("secrets_update", Some(user_id), e))?;
+                (entry.name, Some(entry.folder))
+            } else {
+                (input.name.clone(), input.folder.clone())
+            };
+
         let add_tags = input.add_tags.unwrap_or_default();
         let remove_tags = input.remove_tags.unwrap_or_default();
-        let meta = input.meta.unwrap_or_default();
+        let mut meta = input.meta.unwrap_or_default();
+        if let Some(obj) = input.meta_obj {
+            meta.extend(map_to_kv_strings(obj));
+        }
         let remove_meta = input.remove_meta.unwrap_or_default();
-        let secrets = input.secrets.unwrap_or_default();
+        let mut secrets = input.secrets.unwrap_or_default();
+        if let Some(obj) = input.secrets_obj {
+            secrets.extend(map_to_kv_strings(obj));
+        }
         let remove_secrets = input.remove_secrets.unwrap_or_default();
 
         let result = svc_update(
             &self.pool,
             UpdateParams {
-                namespace: &input.namespace,
-                kind: &input.kind,
-                name: &input.name,
+                name: &resolved_name,
+                folder: resolved_folder.as_deref(),
+                notes: input.notes.as_deref(),
                 add_tags: &add_tags,
                 remove_tags: &remove_tags,
                 meta_entries: &meta,
@@ -567,9 +739,7 @@ impl SecretsService {
         tracing::info!(
             tool = "secrets_update",
             ?user_id,
-            namespace = %input.namespace,
-            kind = %input.kind,
-            name = %input.name,
+            name = %resolved_name,
             elapsed_ms = t.elapsed().as_millis(),
             "tool call ok",
         );
@@ -578,8 +748,9 @@ impl SecretsService {
     }
 
     #[tool(
-        description = "Delete one entry (specify namespace+kind+name) or bulk delete all \
-        entries matching namespace+kind. Use dry_run=true to preview.",
+        description = "Delete one entry by name (or id), or bulk delete entries matching folder \
+        and/or type. Use dry_run=true to preview. \
+        At least one of id, name, folder, or type must be provided.",
         annotations(title = "Delete Secret Entry", destructive_hint = true)
     )]
     async fn secrets_delete(
@@ -589,22 +760,49 @@ impl SecretsService {
     ) -> Result<CallToolResult, rmcp::ErrorData> {
         let t = Instant::now();
         let user_id = Self::user_id_from_ctx(&ctx)?;
+
+        // Safety: require at least one filter.
+        if input.id.is_none()
+            && input.name.is_none()
+            && input.folder.is_none()
+            && input.entry_type.is_none()
+        {
+            return Err(rmcp::ErrorData::invalid_request(
+                "At least one of id, name, folder, or type must be provided.",
+                None,
+            ));
+        }
+
         tracing::info!(
             tool = "secrets_delete",
             ?user_id,
-            namespace = %input.namespace,
-            kind = input.kind.as_deref(),
+            id = ?input.id,
             name = input.name.as_deref(),
+            folder = input.folder.as_deref(),
+            entry_type = input.entry_type.as_deref(),
             dry_run = input.dry_run.unwrap_or(false),
             "tool call start",
         );
 
+        // When id is provided, resolve to name+folder for the single-entry delete path.
+        let (effective_name, effective_folder): (Option<String>, Option<String>) =
+            if let Some(ref id_str) = input.id {
+                let eid = parse_uuid(id_str)?;
+                let uid = user_id;
+                let entry = resolve_entry_by_id(&self.pool, eid, uid)
+                    .await
+                    .map_err(|e| mcp_err_internal_logged("secrets_delete", uid, e))?;
+                (Some(entry.name), Some(entry.folder))
+            } else {
+                (input.name.clone(), input.folder.clone())
+            };
+
         let result = svc_delete(
             &self.pool,
             DeleteParams {
-                namespace: &input.namespace,
-                kind: input.kind.as_deref(),
-                name: input.name.as_deref(),
+                name: effective_name.as_deref(),
+                folder: effective_folder.as_deref(),
+                entry_type: input.entry_type.as_deref(),
                 dry_run: input.dry_run.unwrap_or(false),
                 user_id,
             },
@@ -615,7 +813,6 @@ impl SecretsService {
         tracing::info!(
             tool = "secrets_delete",
             ?user_id,
-            namespace = %input.namespace,
             elapsed_ms = t.elapsed().as_millis(),
             "tool call ok",
         );
@@ -625,7 +822,7 @@ impl SecretsService {
 
     #[tool(
         description = "View change history for an entry. Returns a list of versions with \
-        actions and timestamps.",
+        actions and timestamps. Optionally pass 'id' (from secrets_find) to target directly.",
         annotations(
             title = "View Secret History",
             read_only_hint = true,
@@ -642,17 +839,26 @@ impl SecretsService {
         tracing::info!(
             tool = "secrets_history",
             ?user_id,
-            namespace = %input.namespace,
-            kind = %input.kind,
             name = %input.name,
+            id = ?input.id,
             "tool call start",
         );
 
+        let (resolved_name, resolved_folder): (String, Option<String>) =
+            if let Some(ref id_str) = input.id {
+                let eid = parse_uuid(id_str)?;
+                let entry = resolve_entry_by_id(&self.pool, eid, user_id)
+                    .await
+                    .map_err(|e| mcp_err_internal_logged("secrets_history", user_id, e))?;
+                (entry.name, Some(entry.folder))
+            } else {
+                (input.name.clone(), input.folder.clone())
+            };
+
         let result = svc_history(
             &self.pool,
-            &input.namespace,
-            &input.kind,
-            &input.name,
+            &resolved_name,
+            resolved_folder.as_deref(),
             input.limit.unwrap_or(20),
             user_id,
         )
@@ -671,7 +877,8 @@ impl SecretsService {
 
     #[tool(
         description = "Rollback an entry to a previous version. Requires X-Encryption-Key header. \
-        Omit to_version to restore the most recent snapshot.",
+        Omit to_version to restore the most recent snapshot. \
+        Optionally pass 'id' (from secrets_find) to target directly.",
         annotations(title = "Rollback Secret Entry", destructive_hint = true)
     )]
     async fn secrets_rollback(
@@ -684,18 +891,27 @@ impl SecretsService {
         tracing::info!(
             tool = "secrets_rollback",
             ?user_id,
-            namespace = %input.namespace,
-            kind = %input.kind,
             name = %input.name,
+            id = ?input.id,
             to_version = input.to_version,
             "tool call start",
         );
 
+        let (resolved_name, resolved_folder): (String, Option<String>) =
+            if let Some(ref id_str) = input.id {
+                let eid = parse_uuid(id_str)?;
+                let entry = resolve_entry_by_id(&self.pool, eid, Some(user_id))
+                    .await
+                    .map_err(|e| mcp_err_internal_logged("secrets_rollback", Some(user_id), e))?;
+                (entry.name, Some(entry.folder))
+            } else {
+                (input.name.clone(), input.folder.clone())
+            };
+
         let result = svc_rollback(
             &self.pool,
-            &input.namespace,
-            &input.kind,
-            &input.name,
+            &resolved_name,
+            resolved_folder.as_deref(),
             input.to_version,
             &user_key,
             Some(user_id),
@@ -734,8 +950,8 @@ impl SecretsService {
         tracing::info!(
             tool = "secrets_export",
             ?user_id,
-            namespace = input.namespace.as_deref(),
-            kind = input.kind.as_deref(),
+            folder = input.folder.as_deref(),
+            entry_type = input.entry_type.as_deref(),
             format,
             "tool call start",
         );
@@ -743,8 +959,8 @@ impl SecretsService {
         let data = svc_export(
             &self.pool,
             ExportParams {
-                namespace: input.namespace.as_deref(),
-                kind: input.kind.as_deref(),
+                folder: input.folder.as_deref(),
+                entry_type: input.entry_type.as_deref(),
                 name: input.name.as_deref(),
                 tags: &tags,
                 query: input.query.as_deref(),
@@ -785,7 +1001,10 @@ impl SecretsService {
     #[tool(
         description = "Build the environment variable map from entry secrets with decrypted \
         plaintext values. Requires X-Encryption-Key header. \
-        Returns a JSON object of VAR_NAME -> plaintext_value ready for injection.",
+        Returns a JSON object of VAR_NAME -> plaintext_value ready for injection. \
+        Variable names follow the pattern UPPER(entry_name)_UPPER(field_name), \
+        with hyphens and dots replaced by underscores. \
+        Example: entry 'aliyun', field 'access_key_id' → ALIYUN_ACCESS_KEY_ID.",
         annotations(title = "Build Env Map", read_only_hint = true, idempotent_hint = true)
     )]
     async fn secrets_env_map(
@@ -800,16 +1019,16 @@ impl SecretsService {
         tracing::info!(
             tool = "secrets_env_map",
             ?user_id,
-            namespace = input.namespace.as_deref(),
-            kind = input.kind.as_deref(),
+            folder = input.folder.as_deref(),
+            entry_type = input.entry_type.as_deref(),
             prefix = input.prefix.as_deref().unwrap_or(""),
             "tool call start",
         );
 
         let env_map = secrets_core::service::env_map::build_env_map(
             &self.pool,
-            input.namespace.as_deref(),
-            input.kind.as_deref(),
+            input.folder.as_deref(),
+            input.entry_type.as_deref(),
             input.name.as_deref(),
             &tags,
             &only_fields,
@@ -831,6 +1050,67 @@ impl SecretsService {
         let json = serde_json::to_string_pretty(&env_map).unwrap_or_default();
         Ok(CallToolResult::success(vec![Content::text(json)]))
     }
+
+    #[tool(
+        description = "Get an overview of the secrets store: counts of entries per folder and \
+        per type. Requires Bearer API key. Useful for exploring the store structure.",
+        annotations(
+            title = "Secrets Overview",
+            read_only_hint = true,
+            idempotent_hint = true
+        )
+    )]
+    async fn secrets_overview(
+        &self,
+        Parameters(_input): Parameters<OverviewInput>,
+        ctx: RequestContext<RoleServer>,
+    ) -> Result<CallToolResult, rmcp::ErrorData> {
+        let t = Instant::now();
+        let user_id = Self::require_user_id(&ctx)?;
+        tracing::info!(tool = "secrets_overview", ?user_id, "tool call start");
+
+        #[derive(sqlx::FromRow)]
+        struct CountRow {
+            name: String,
+            count: i64,
+        }
+
+        let folder_rows: Vec<CountRow> = sqlx::query_as(
+            "SELECT folder AS name, COUNT(*) AS count FROM entries \
+             WHERE user_id = $1 GROUP BY folder ORDER BY folder",
+        )
+        .bind(user_id)
+        .fetch_all(&*self.pool)
+        .await
+        .map_err(|e| mcp_err_internal_logged("secrets_overview", Some(user_id), e))?;
+
+        let type_rows: Vec<CountRow> = sqlx::query_as(
+            "SELECT type AS name, COUNT(*) AS count FROM entries \
+             WHERE user_id = $1 GROUP BY type ORDER BY type",
+        )
+        .bind(user_id)
+        .fetch_all(&*self.pool)
+        .await
+        .map_err(|e| mcp_err_internal_logged("secrets_overview", Some(user_id), e))?;
+
+        let total: i64 = folder_rows.iter().map(|r| r.count).sum();
+
+        let result = serde_json::json!({
+            "total": total,
+            "folders": folder_rows.iter().map(|r| serde_json::json!({"name": r.name, "count": r.count})).collect::<Vec<_>>(),
+            "types": type_rows.iter().map(|r| serde_json::json!({"name": r.name, "count": r.count})).collect::<Vec<_>>(),
+        });
+
+        tracing::info!(
+            tool = "secrets_overview",
+            ?user_id,
+            total,
+            elapsed_ms = t.elapsed().as_millis(),
+            "tool call ok",
+        );
+        let json = serde_json::to_string_pretty(&result).unwrap_or_default();
+        Ok(CallToolResult::success(vec![Content::text(json)]))
+    }
 }
 
 // ── ServerHandler ─────────────────────────────────────────────────────────────
@@ -847,11 +1127,11 @@ impl ServerHandler for SecretsService {
         info.protocol_version = ProtocolVersion::V_2025_06_18;
         info.instructions = Some(
             "Manage cross-device secrets and configuration securely. \
-             Data is encrypted with your passphrase-derived key. \
-             Include your 64-char hex key in the X-Encryption-Key header for all read/write operations. \
-             Use secrets_search to discover entries (Bearer token required; encryption key not needed), \
-             secrets_get to decrypt secret values, \
-             and secrets_add/secrets_update to write encrypted secrets."
+             Use secrets_find to discover entries by folder, name, type, tags, or query \
+             (query also searches metadata values). \
+             Use secrets_get with the entry id (from secrets_find) to decrypt secret values. \
+             Use secrets_add / secrets_update to write entries. \
+             Use secrets_overview for a quick count of entries per folder and type."
                 .to_string(),
         );
         info

crates/secrets-mcp/src/web.rs

@@ -8,9 +8,10 @@ use axum::{
     extract::{ConnectInfo, Path, Query, State},
     http::{HeaderMap, StatusCode, header},
     response::{Html, IntoResponse, Redirect, Response},
-    routing::{get, post},
+    routing::{get, patch, post},
 };
 use serde::{Deserialize, Serialize};
+use serde_json::json;
 use tower_sessions::Session;
 use uuid::Uuid;
 
@@ -19,6 +20,9 @@ use secrets_core::crypto::hex;
 use secrets_core::service::{
     api_key::{ensure_api_key, regenerate_api_key},
     audit_log::list_for_user,
+    delete::delete_by_id,
+    search::{SearchParams, count_entries, list_entries},
+    update::{UpdateEntryFieldsByIdParams, update_fields_by_id},
     user::{
         OAuthProfile, bind_oauth_account, find_or_create_user, get_user_by_id,
         unbind_oauth_account, update_user_key_setup,
@@ -78,6 +82,44 @@ struct AuditEntryView {
     detail: String,
 }
 
+#[derive(Template)]
+#[template(path = "entries.html")]
+struct EntriesPageTemplate {
+    user_name: String,
+    user_email: String,
+    entries: Vec<EntryListItemView>,
+    total_count: i64,
+    shown_count: usize,
+    limit: u32,
+    filter_folder: String,
+    filter_type: String,
+    version: &'static str,
+}
+
+/// Non-sensitive fields only (no `secrets` / ciphertext).
+struct EntryListItemView {
+    id: String,
+    folder: String,
+    entry_type: String,
+    name: String,
+    notes: String,
+    tags: String,
+    metadata: String,
+    /// RFC3339 UTC for `<time datetime>`; localized in entries.html.
+    updated_at_iso: String,
+}
+
+/// Cap for HTML list (avoids loading unbounded rows into memory).
+const ENTRIES_PAGE_LIMIT: u32 = 5_000;
+
+#[derive(Deserialize)]
+struct EntriesQuery {
+    folder: Option<String>,
+    /// URL query key is `type` (maps to DB column `entries.type`).
+    #[serde(rename = "type")]
+    entry_type: Option<String>,
+}
+
 // ── App state helpers ─────────────────────────────────────────────────────────
 
 fn google_cfg(state: &AppState) -> Option<&OAuthConfig> {
@@ -149,6 +191,7 @@ pub fn web_router() -> Router<AppState> {
         .route("/auth/google/callback", get(auth_google_callback))
         .route("/auth/logout", post(auth_logout))
         .route("/dashboard", get(dashboard))
+        .route("/entries", get(entries_page))
         .route("/audit", get(audit_page))
         .route("/account/bind/google", get(account_bind_google))
         .route(
@@ -160,6 +203,10 @@ pub fn web_router() -> Router<AppState> {
         .route("/api/key-setup", post(api_key_setup))
         .route("/api/apikey", get(api_apikey_get))
         .route("/api/apikey/regenerate", post(api_apikey_regenerate))
+        .route(
+            "/api/entries/{id}",
+            patch(api_entry_patch).delete(api_entry_delete),
+        )
 }
 
 fn text_asset_response(content: &'static str, content_type: &'static str) -> Response {
@@ -478,6 +525,89 @@ async fn dashboard(
     render_template(tmpl)
 }
 
+async fn entries_page(
+    State(state): State<AppState>,
+    session: Session,
+    Query(q): Query<EntriesQuery>,
+) -> Result<Response, StatusCode> {
+    let Some(user_id) = current_user_id(&session).await else {
+        return Ok(Redirect::to("/login").into_response());
+    };
+
+    let user = match get_user_by_id(&state.pool, user_id).await.map_err(|e| {
+        tracing::error!(error = %e, %user_id, "failed to load user for entries page");
+        StatusCode::INTERNAL_SERVER_ERROR
+    })? {
+        Some(u) => u,
+        None => return Ok(Redirect::to("/login").into_response()),
+    };
+
+    let folder_filter = q
+        .folder
+        .as_ref()
+        .map(|s| s.trim())
+        .filter(|s| !s.is_empty())
+        .map(|s| s.to_string());
+    let type_filter = q
+        .entry_type
+        .as_ref()
+        .map(|s| s.trim())
+        .filter(|s| !s.is_empty())
+        .map(|s| s.to_string());
+
+    let params = SearchParams {
+        folder: folder_filter.as_deref(),
+        entry_type: type_filter.as_deref(),
+        name: None,
+        tags: &[],
+        query: None,
+        sort: "updated",
+        limit: ENTRIES_PAGE_LIMIT,
+        offset: 0,
+        user_id: Some(user_id),
+    };
+
+    let total_count = count_entries(&state.pool, &params).await.map_err(|e| {
+        tracing::error!(error = %e, "failed to count entries for web");
+        StatusCode::INTERNAL_SERVER_ERROR
+    })?;
+
+    let rows = list_entries(&state.pool, params).await.map_err(|e| {
+        tracing::error!(error = %e, "failed to load entries list for web");
+        StatusCode::INTERNAL_SERVER_ERROR
+    })?;
+    let shown_count = rows.len();
+
+    let entries = rows
+        .into_iter()
+        .map(|e| EntryListItemView {
+            id: e.id.to_string(),
+            folder: e.folder,
+            entry_type: e.entry_type,
+            name: e.name,
+            notes: e.notes,
+            tags: e.tags.join(", "),
+            metadata: serde_json::to_string_pretty(&e.metadata)
+                .unwrap_or_else(|_| "{}".to_string()),
+            updated_at_iso: e.updated_at.to_rfc3339_opts(SecondsFormat::Secs, true),
+        })
+        .collect();
+
+    let tmpl = EntriesPageTemplate {
+        user_name: user.name.clone(),
+        user_email: user.email.clone().unwrap_or_default(),
+        entries,
+        total_count,
+        shown_count,
+        limit: ENTRIES_PAGE_LIMIT,
+        filter_folder: folder_filter.unwrap_or_default(),
+        filter_type: type_filter.unwrap_or_default(),
+        version: env!("CARGO_PKG_VERSION"),
+    };
+
+    render_template(tmpl)
+}
+
 async fn audit_page(
     State(state): State<AppState>,
     session: Session,
@@ -506,7 +636,7 @@ async fn audit_page(
         .map(|row| AuditEntryView {
             created_at_iso: row.created_at.to_rfc3339_opts(SecondsFormat::Secs, true),
             action: row.action,
-            target: format_audit_target(&row.namespace, &row.kind, &row.name),
+            target: format_audit_target(&row.folder, &row.entry_type, &row.name),
             detail: serde_json::to_string_pretty(&row.detail).unwrap_or_else(|_| "{}".to_string()),
         })
         .collect();
@@ -751,6 +881,125 @@ async fn api_apikey_regenerate(
     Ok(Json(ApiKeyResponse { api_key }))
 }
 
+// ── Entry management (Web UI, non-sensitive fields only) ───────────────────────
+
+#[derive(Deserialize)]
+struct EntryPatchBody {
+    folder: String,
+    #[serde(rename = "type")]
+    entry_type: String,
+    name: String,
+    notes: String,
+    tags: Vec<String>,
+    metadata: serde_json::Value,
+}
+
+type EntryApiError = (StatusCode, Json<serde_json::Value>);
+
+fn map_entry_mutation_err(e: anyhow::Error) -> EntryApiError {
+    let msg = e.to_string();
+    if msg.contains("Entry not found") {
+        return (
+            StatusCode::NOT_FOUND,
+            Json(json!({ "error": "条目不存在或无权访问" })),
+        );
+    }
+    if msg.contains("already exists") {
+        return (
+            StatusCode::CONFLICT,
+            Json(json!({ "error": "该账号下已存在相同 folder + name 的条目" })),
+        );
+    }
+    if msg.contains("Concurrent modification") {
+        return (
+            StatusCode::CONFLICT,
+            Json(json!({ "error": "条目已被修改，请刷新后重试" })),
+        );
+    }
+    if msg.contains("must be at most") {
+        return (StatusCode::BAD_REQUEST, Json(json!({ "error": msg })));
+    }
+    tracing::error!(error = %e, "entry mutation failed");
+    (
+        StatusCode::INTERNAL_SERVER_ERROR,
+        Json(json!({ "error": "操作失败，请稍后重试" })),
+    )
+}
+
+async fn api_entry_patch(
+    State(state): State<AppState>,
+    session: Session,
+    Path(entry_id): Path<Uuid>,
+    Json(body): Json<EntryPatchBody>,
+) -> Result<Json<serde_json::Value>, EntryApiError> {
+    let user_id = current_user_id(&session)
+        .await
+        .ok_or((StatusCode::UNAUTHORIZED, Json(json!({ "error": "未登录" }))))?;
+
+    let folder = body.folder.trim();
+    let entry_type = body.entry_type.trim();
+    let name = body.name.trim();
+    let notes = body.notes.trim();
+
+    if name.is_empty() {
+        return Err((
+            StatusCode::BAD_REQUEST,
+            Json(json!({ "error": "name 不能为空" })),
+        ));
+    }
+
+    let tags: Vec<String> = body
+        .tags
+        .into_iter()
+        .map(|t| t.trim().to_string())
+        .filter(|t| !t.is_empty())
+        .collect();
+
+    if !body.metadata.is_object() {
+        return Err((
+            StatusCode::BAD_REQUEST,
+            Json(json!({ "error": "metadata 必须是 JSON 对象" })),
+        ));
+    }
+
+    update_fields_by_id(
+        &state.pool,
+        entry_id,
+        user_id,
+        UpdateEntryFieldsByIdParams {
+            folder,
+            entry_type,
+            name,
+            notes,
+            tags: &tags,
+            metadata: &body.metadata,
+        },
+    )
+    .await
+    .map_err(map_entry_mutation_err)?;
+
+    Ok(Json(json!({ "ok": true })))
+}
+
+async fn api_entry_delete(
+    State(state): State<AppState>,
+    session: Session,
+    Path(entry_id): Path<Uuid>,
+) -> Result<Json<serde_json::Value>, EntryApiError> {
+    let user_id = current_user_id(&session)
+        .await
+        .ok_or((StatusCode::UNAUTHORIZED, Json(json!({ "error": "未登录" }))))?;
+
+    let result = delete_by_id(&state.pool, entry_id, user_id)
+        .await
+        .map_err(map_entry_mutation_err)?;
+
+    Ok(Json(json!({
+        "ok": true,
+        "migrated": result.migrated,
+    })))
+}
+
 // ── OAuth / Well-known ────────────────────────────────────────────────────────
 
 /// RFC 9728 — OAuth 2.0 Protected Resource Metadata.
@@ -783,11 +1032,15 @@ fn render_template<T: Template>(tmpl: T) -> Result<Response, StatusCode> {
     Ok(Html(html).into_response())
 }
 
-fn format_audit_target(namespace: &str, kind: &str, name: &str) -> String {
-    // Auth events reuse kind/name as a provider-scoped target, not an entry identity.
-    if namespace == "auth" {
-        format!("{}/{}", kind, name)
+fn format_audit_target(folder: &str, entry_type: &str, name: &str) -> String {
+    // Auth events (folder="auth") use entry_type/name as provider-scoped target.
+    if folder == "auth" {
+        format!("{}/{}", entry_type, name)
+    } else if !folder.is_empty() && !entry_type.is_empty() {
+        format!("[{}/{}] {}", folder, entry_type, name)
+    } else if !folder.is_empty() {
+        format!("[{}] {}", folder, name)
     } else {
-        format!("[{}/{}] {}", namespace, kind, name)
+        name.to_string()
     }
 }

crates/secrets-mcp/templates/audit.html

@@ -92,6 +92,7 @@
     <a href="/dashboard" class="sidebar-logo"><span>secrets</span></a>
     <nav class="sidebar-menu">
         <a href="/dashboard" class="sidebar-link">MCP</a>
+        <a href="/entries" class="sidebar-link">条目</a>
         <a href="/audit" class="sidebar-link active">审计</a>
     </nav>
 </aside>

crates/secrets-mcp/templates/dashboard.html

@@ -174,6 +174,7 @@
     <a href="/dashboard" class="sidebar-logo"><span>secrets</span></a>
     <nav class="sidebar-menu">
         <a href="/dashboard" class="sidebar-link active">MCP</a>
+        <a href="/entries" class="sidebar-link">条目</a>
         <a href="/audit" class="sidebar-link">审计</a>
     </nav>
 </aside>

crates/secrets-mcp/templates/entries.html

@@ -0,0 +1,390 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <link rel="icon" href="/favicon.svg?v={{ version }}" type="image/svg+xml">
+    <title>Secrets — 条目</title>
+    <style>
+        *, *::before, *::after { box-sizing: border-box; margin: 0; padding: 0; }
+        @import url('https://fonts.googleapis.com/css2?family=JetBrains+Mono:wght@400;600&family=Inter:wght@400;500;600&display=swap');
+        :root {
+            --bg: #0d1117; --surface: #161b22; --surface2: #21262d;
+            --border: #30363d; --text: #e6edf3; --text-muted: #8b949e;
+            --accent: #58a6ff; --accent-hover: #79b8ff;
+        }
+        body { background: var(--bg); color: var(--text); font-family: 'Inter', sans-serif; min-height: 100vh; }
+        .layout { display: flex; min-height: 100vh; }
+        .sidebar {
+            width: 220px; flex-shrink: 0; background: var(--surface); border-right: 1px solid var(--border);
+            padding: 24px 16px; display: flex; flex-direction: column; gap: 20px;
+        }
+        .sidebar-logo { font-family: 'JetBrains Mono', monospace; font-size: 16px; font-weight: 600;
+                        color: var(--text); text-decoration: none; padding: 0 10px; }
+        .sidebar-logo span { color: var(--accent); }
+        .sidebar-menu { display: flex; flex-direction: column; gap: 6px; }
+        .sidebar-link {
+            padding: 10px 12px; border-radius: 8px; color: var(--text-muted); text-decoration: none;
+            border: 1px solid transparent; font-size: 13px; font-weight: 500;
+        }
+        .sidebar-link:hover { background: var(--surface2); color: var(--text); }
+        .sidebar-link.active {
+            background: rgba(88,166,255,0.12); color: var(--text); border-color: rgba(88,166,255,0.35);
+        }
+        .content-shell { flex: 1; min-width: 0; display: flex; flex-direction: column; }
+        .topbar {
+            background: var(--surface); border-bottom: 1px solid var(--border); padding: 0 24px;
+            display: flex; align-items: center; gap: 12px; min-height: 52px;
+        }
+        .topbar-spacer { flex: 1; }
+        .nav-user { font-size: 13px; color: var(--text-muted); }
+        .btn-sign-out {
+            padding: 5px 12px; border-radius: 6px; border: 1px solid var(--border);
+            background: none; color: var(--text); font-size: 12px; text-decoration: none; cursor: pointer;
+        }
+        .btn-sign-out:hover { background: var(--surface2); }
+        .main { padding: 32px 24px 40px; flex: 1; }
+        .card { background: var(--surface); border: 1px solid var(--border); border-radius: 12px;
+                padding: 24px; width: 100%; max-width: 1280px; margin: 0 auto; }
+        .card-title { font-size: 20px; font-weight: 600; margin-bottom: 8px; }
+        .card-subtitle { color: var(--text-muted); font-size: 13px; margin-bottom: 20px; }
+        .filter-bar {
+            display: flex; flex-wrap: wrap; align-items: flex-end; gap: 12px 16px;
+            margin-bottom: 20px; padding: 16px; background: var(--bg); border: 1px solid var(--border);
+            border-radius: 10px;
+        }
+        .filter-field { display: flex; flex-direction: column; gap: 6px; min-width: 140px; flex: 1; }
+        .filter-field label { font-size: 12px; color: var(--text-muted); font-weight: 500; }
+        .filter-field input {
+            background: var(--surface); border: 1px solid var(--border); border-radius: 6px;
+            color: var(--text); padding: 8px 10px; font-size: 13px; font-family: 'JetBrains Mono', monospace;
+            outline: none; width: 100%;
+        }
+        .filter-field input:focus { border-color: var(--accent); }
+        .filter-actions { display: flex; flex-wrap: wrap; align-items: center; gap: 8px; }
+        .btn-filter {
+            padding: 8px 16px; border-radius: 6px; border: none; background: var(--accent); color: #0d1117;
+            font-size: 13px; font-weight: 600; cursor: pointer;
+        }
+        .btn-filter:hover { background: var(--accent-hover); }
+        .btn-clear {
+            padding: 8px 14px; border-radius: 6px; border: 1px solid var(--border); background: transparent;
+            color: var(--text-muted); font-size: 13px; text-decoration: none; cursor: pointer;
+        }
+        .btn-clear:hover { background: var(--surface2); color: var(--text); }
+        .empty { color: var(--text-muted); font-size: 14px; padding: 20px 0; }
+        .table-wrap { overflow-x: auto; }
+        table { width: 100%; border-collapse: collapse; min-width: 720px; }
+        th, td { text-align: left; vertical-align: top; padding: 12px 10px; border-top: 1px solid var(--border); }
+        th { color: var(--text-muted); font-size: 12px; font-weight: 600; white-space: nowrap; }
+        td { font-size: 13px; }
+        .mono { font-family: 'JetBrains Mono', monospace; }
+        .cell-notes, .cell-meta {
+            max-width: 280px; word-break: break-word;
+        }
+        .notes-scroll {
+            max-height: 160px;
+            overflow: auto;
+            white-space: pre-wrap;
+            word-break: break-word;
+            padding: 8px;
+            background: var(--bg);
+            border: 1px solid var(--border);
+            border-radius: 8px;
+            font-size: 12px;
+        }
+        .detail {
+            background: var(--bg); border: 1px solid var(--border); border-radius: 8px;
+            padding: 10px; white-space: pre-wrap; word-break: break-word; font-size: 12px;
+            max-width: 320px; max-height: 160px; overflow: auto;
+        }
+        .col-actions { white-space: nowrap; }
+        .row-actions { display: flex; flex-wrap: wrap; gap: 6px; }
+        .btn-row {
+            padding: 4px 10px; border-radius: 6px; font-size: 12px; cursor: pointer;
+            border: 1px solid var(--border); background: var(--surface2); color: var(--text-muted);
+            font-family: inherit;
+        }
+        .btn-row:hover { color: var(--text); border-color: var(--text-muted); }
+        .btn-row.danger:hover { border-color: #f85149; color: #f85149; }
+        .modal-overlay {
+            position: fixed; inset: 0; background: rgba(1, 4, 9, 0.65); z-index: 200;
+            display: flex; align-items: center; justify-content: center; padding: 16px;
+        }
+        .modal-overlay[hidden] { display: none !important; }
+        .modal {
+            background: var(--surface); border: 1px solid var(--border); border-radius: 12px;
+            padding: 22px; width: 100%; max-width: 520px; max-height: 90vh; overflow: auto;
+            box-shadow: 0 16px 48px rgba(0,0,0,0.45);
+        }
+        .modal-title { font-size: 16px; font-weight: 600; margin-bottom: 14px; }
+        .modal-field { margin-bottom: 12px; }
+        .modal-field label { display: block; font-size: 12px; color: var(--text-muted); margin-bottom: 5px; }
+        .modal-field input, .modal-field textarea {
+            width: 100%; background: var(--bg); border: 1px solid var(--border); border-radius: 6px;
+            color: var(--text); padding: 8px 10px; font-size: 13px; font-family: 'JetBrains Mono', monospace;
+            outline: none;
+        }
+        .modal-field textarea { min-height: 72px; resize: vertical; }
+        .modal-field textarea.metadata-edit { min-height: 140px; }
+        .modal-error { color: #f85149; font-size: 12px; margin-bottom: 10px; display: none; }
+        .modal-error.visible { display: block; }
+        .modal-footer { display: flex; flex-wrap: wrap; gap: 8px; justify-content: flex-end; margin-top: 16px; }
+        .btn-modal { padding: 8px 16px; border-radius: 6px; font-size: 13px; cursor: pointer; font-family: inherit; border: 1px solid var(--border); background: transparent; color: var(--text); }
+        .btn-modal.primary { background: var(--accent); color: #0d1117; border-color: transparent; font-weight: 600; }
+        .btn-modal.primary:hover { background: var(--accent-hover); }
+        .btn-modal.danger { border-color: #f85149; color: #f85149; }
+        @media (max-width: 900px) {
+            .layout { flex-direction: column; }
+            .sidebar {
+                width: 100%; border-right: none; border-bottom: 1px solid var(--border);
+                padding: 16px; gap: 14px;
+            }
+            .sidebar-menu { flex-direction: row; flex-wrap: wrap; }
+            .sidebar-link { flex: 1; text-align: center; min-width: 72px; }
+            .main { padding: 20px 12px 28px; }
+            .card { padding: 16px; }
+            .topbar { padding: 12px 16px; flex-wrap: wrap; }
+            table, thead, tbody, th, td, tr { display: block; }
+            thead { display: none; }
+            tr { border-top: 1px solid var(--border); padding: 12px 0; }
+            td { border-top: none; padding: 6px 0; max-width: none; }
+            td::before {
+                display: block; color: var(--text-muted); font-size: 11px;
+                margin-bottom: 4px; text-transform: uppercase;
+            }
+            td.col-updated::before { content: "更新"; }
+            td.col-folder::before { content: "Folder"; }
+            td.col-type::before { content: "Type"; }
+            td.col-name::before { content: "Name"; }
+            td.col-notes::before { content: "Notes"; }
+            td.col-tags::before { content: "Tags"; }
+            td.col-meta::before { content: "Metadata"; }
+            td.col-actions::before { content: "操作"; }
+            .detail { max-width: none; }
+            .notes-scroll { max-width: none; }
+        }
+    </style>
+</head>
+<body>
+<div class="layout">
+<aside class="sidebar">
+    <a href="/dashboard" class="sidebar-logo"><span>secrets</span></a>
+    <nav class="sidebar-menu">
+        <a href="/dashboard" class="sidebar-link">MCP</a>
+        <a href="/entries" class="sidebar-link active">条目</a>
+        <a href="/audit" class="sidebar-link">审计</a>
+    </nav>
+</aside>
+
+<div class="content-shell">
+<div class="topbar">
+    <span class="topbar-spacer"></span>
+    <span class="nav-user">{{ user_name }}{% if !user_email.is_empty() %} · {{ user_email }}{% endif %}</span>
+    <form action="/auth/logout" method="post" style="display:inline">
+        <button type="submit" class="btn-sign-out">退出</button>
+    </form>
+</div>
+
+<main class="main">
+    <section class="card">
+        <div class="card-title">我的条目</div>
+        <div class="card-subtitle">在当前筛选条件下，共 <strong>{{ total_count }}</strong> 条记录；本页显示 <strong>{{ shown_count }}</strong> 条（按更新时间降序，单页最多 {{ limit }} 条）。不含密文字段。时间为浏览器本地时区。</div>
+
+        <form class="filter-bar" method="get" action="/entries">
+            <div class="filter-field">
+                <label for="filter-folder">Folder（精确匹配）</label>
+                <input id="filter-folder" name="folder" type="text" value="{{ filter_folder }}" placeholder="例如 refining" autocomplete="off">
+            </div>
+            <div class="filter-field">
+                <label for="filter-type">Type（精确匹配）</label>
+                <input id="filter-type" name="type" type="text" value="{{ filter_type }}" placeholder="例如 server" autocomplete="off">
+            </div>
+            <div class="filter-actions">
+                <button type="submit" class="btn-filter">筛选</button>
+                <a href="/entries" class="btn-clear">清空</a>
+            </div>
+        </form>
+
+        {% if entries.is_empty() %}
+        <div class="empty">暂无条目。</div>
+        {% else %}
+        <div class="table-wrap">
+        <table>
+            <thead>
+                <tr>
+                    <th>更新</th>
+                    <th>Folder</th>
+                    <th>Type</th>
+                    <th>Name</th>
+                    <th>Notes</th>
+                    <th>Tags</th>
+                    <th>Metadata</th>
+                    <th>操作</th>
+                </tr>
+            </thead>
+            <tbody>
+                {% for entry in entries %}
+                <tr data-entry-id="{{ entry.id }}">
+                    <td class="col-updated mono"><time class="entry-local-time" datetime="{{ entry.updated_at_iso }}">{{ entry.updated_at_iso }}</time></td>
+                    <td class="col-folder mono cell-folder">{{ entry.folder }}</td>
+                    <td class="col-type mono cell-type">{{ entry.entry_type }}</td>
+                    <td class="col-name mono cell-name">{{ entry.name }}</td>
+                    <td class="col-notes cell-notes">{% if !entry.notes.is_empty() %}<div class="notes-scroll cell-notes-val">{{ entry.notes }}</div>{% endif %}</td>
+                    <td class="col-tags mono cell-tags-val">{{ entry.tags }}</td>
+                    <td class="col-meta cell-meta"><pre class="detail cell-meta-val">{{ entry.metadata }}</pre></td>
+                    <td class="col-actions">
+                        <div class="row-actions">
+                            <button type="button" class="btn-row btn-edit">编辑</button>
+                            <button type="button" class="btn-row danger btn-del">删除</button>
+                        </div>
+                    </td>
+                </tr>
+                {% endfor %}
+            </tbody>
+        </table>
+        </div>
+        {% endif %}
+    </section>
+</main>
+</div>
+</div>
+
+<div id="edit-overlay" class="modal-overlay" hidden>
+    <div class="modal" role="dialog" aria-modal="true" aria-labelledby="edit-title">
+        <div class="modal-title" id="edit-title">编辑条目</div>
+        <div id="edit-error" class="modal-error"></div>
+        <div class="modal-field"><label for="edit-folder">Folder</label><input id="edit-folder" type="text" autocomplete="off"></div>
+        <div class="modal-field"><label for="edit-type">Type</label><input id="edit-type" type="text" autocomplete="off"></div>
+        <div class="modal-field"><label for="edit-name">Name</label><input id="edit-name" type="text" autocomplete="off"></div>
+        <div class="modal-field"><label for="edit-notes">Notes</label><textarea id="edit-notes"></textarea></div>
+        <div class="modal-field"><label for="edit-tags">Tags（逗号分隔）</label><input id="edit-tags" type="text" autocomplete="off"></div>
+        <div class="modal-field"><label for="edit-metadata">Metadata（JSON 对象）</label><textarea id="edit-metadata" class="metadata-edit"></textarea></div>
+        <div class="modal-footer">
+            <button type="button" class="btn-modal" id="edit-cancel">取消</button>
+            <button type="button" class="btn-modal primary" id="edit-save">保存</button>
+        </div>
+    </div>
+</div>
+<script>
+(function () {
+  document.querySelectorAll('time.entry-local-time[datetime]').forEach(function (el) {
+    var raw = el.getAttribute('datetime');
+    var d = raw ? new Date(raw) : null;
+    if (d && !isNaN(d.getTime())) {
+      el.textContent = d.toLocaleString(undefined, { dateStyle: 'medium', timeStyle: 'medium' });
+      el.title = raw + ' (UTC)';
+    }
+  });
+
+  var editOverlay = document.getElementById('edit-overlay');
+  var editError = document.getElementById('edit-error');
+  var editFolder = document.getElementById('edit-folder');
+  var editType = document.getElementById('edit-type');
+  var editName = document.getElementById('edit-name');
+  var editNotes = document.getElementById('edit-notes');
+  var editTags = document.getElementById('edit-tags');
+  var editMetadata = document.getElementById('edit-metadata');
+  var currentEntryId = null;
+
+  function showEditErr(msg) {
+    editError.textContent = msg || '';
+    editError.classList.toggle('visible', !!msg);
+  }
+
+  function openEdit(tr) {
+    var id = tr.getAttribute('data-entry-id');
+    if (!id) return;
+    currentEntryId = id;
+    showEditErr('');
+    editFolder.value = tr.querySelector('.cell-folder') ? tr.querySelector('.cell-folder').textContent.trim() : '';
+    editType.value = tr.querySelector('.cell-type') ? tr.querySelector('.cell-type').textContent.trim() : '';
+    editName.value = tr.querySelector('.cell-name') ? tr.querySelector('.cell-name').textContent.trim() : '';
+    editNotes.value = tr.querySelector('.cell-notes-val') ? tr.querySelector('.cell-notes-val').textContent : '';
+    var tagsText = tr.querySelector('.cell-tags-val') ? tr.querySelector('.cell-tags-val').textContent.trim() : '';
+    editTags.value = tagsText;
+    var metaPre = tr.querySelector('.cell-meta-val');
+    editMetadata.value = metaPre ? metaPre.textContent : '{}';
+    editOverlay.hidden = false;
+  }
+
+  function closeEdit() {
+    editOverlay.hidden = true;
+    currentEntryId = null;
+    showEditErr('');
+  }
+
+  document.getElementById('edit-cancel').addEventListener('click', closeEdit);
+  editOverlay.addEventListener('click', function (e) {
+    if (e.target === editOverlay) closeEdit();
+  });
+
+  document.getElementById('edit-save').addEventListener('click', function () {
+    if (!currentEntryId) return;
+    var meta;
+    try {
+      meta = JSON.parse(editMetadata.value);
+    } catch (err) {
+      showEditErr('Metadata 不是合法 JSON');
+      return;
+    }
+    if (meta === null || typeof meta !== 'object' || Array.isArray(meta)) {
+      showEditErr('Metadata 必须是 JSON 对象');
+      return;
+    }
+    var tags = editTags.value.split(',').map(function (s) { return s.trim(); }).filter(Boolean);
+    var body = {
+      folder: editFolder.value,
+      type: editType.value,
+      name: editName.value.trim(),
+      notes: editNotes.value,
+      tags: tags,
+      metadata: meta
+    };
+    showEditErr('');
+    fetch('/api/entries/' + encodeURIComponent(currentEntryId), {
+      method: 'PATCH',
+      headers: { 'Content-Type': 'application/json' },
+      credentials: 'same-origin',
+      body: JSON.stringify(body)
+    }).then(function (r) {
+      return r.json().then(function (data) {
+        if (!r.ok) throw new Error(data.error || ('HTTP ' + r.status));
+        return data;
+      });
+    }).then(function () {
+      closeEdit();
+      window.location.reload();
+    }).catch(function (e) {
+      showEditErr(e.message || String(e));
+    });
+  });
+
+  document.querySelectorAll('tr[data-entry-id]').forEach(function (tr) {
+    tr.querySelector('.btn-edit').addEventListener('click', function () { openEdit(tr); });
+    tr.querySelector('.btn-del').addEventListener('click', function () {
+      var id = tr.getAttribute('data-entry-id');
+      var nameEl = tr.querySelector('.cell-name');
+      var name = nameEl ? nameEl.textContent.trim() : '';
+      if (!id) return;
+      if (!confirm('确定删除条目「' + name + '」？')) return;
+      fetch('/api/entries/' + encodeURIComponent(id), { method: 'DELETE', credentials: 'same-origin' })
+        .then(function (r) {
+          return r.json().then(function (data) {
+            if (!r.ok) throw new Error(data.error || ('HTTP ' + r.status));
+            return data;
+          });
+        })
+        .then(function (data) {
+          if (data && Array.isArray(data.migrated) && data.migrated.length > 0) {
+            alert('已自动迁移共享 key 引用：' + data.migrated.length + ' 个条目完成重定向。');
+          }
+          window.location.reload();
+        })
+        .catch(function (e) { alert(e.message || String(e)); });
+    });
+  });
+})();
+</script>
+</body>
+</html>

deploy/.env.example

@@ -3,7 +3,13 @@
 
 # ─── 数据库 ───────────────────────────────────────────────────────────
 # Web 会话（tower-sessions）与业务数据共用此库；启动时会自动 migrate 会话表，无需额外环境变量。
-SECRETS_DATABASE_URL=postgres://postgres:PASSWORD@HOST:PORT/secrets-mcp
+SECRETS_DATABASE_URL=postgres://postgres:PASSWORD@db.refining.ltd:5432/secrets-mcp
+# 强烈建议生产使用 verify-full（至少 verify-ca）
+SECRETS_DATABASE_SSL_MODE=verify-full
+# 私有 CA 或自建链路时填写 CA 根证书路径；使用公共受信 CA 可留空
+# SECRETS_DATABASE_SSL_ROOT_CERT=/etc/secrets/pg-ca.crt
+# 当设为 prod/production 时，服务会拒绝弱 TLS 模式（prefer/disable/allow/require）
+SECRETS_ENV=production
 
 # ─── 服务地址 ─────────────────────────────────────────────────────────
 # 内网监听地址（Cloudflare / Nginx 反代时填内网端口）

deploy/postgres-tls-hardening.md

@@ -0,0 +1,92 @@
+# PostgreSQL TLS Hardening Runbook
+
+This runbook applies to:
+
+- PostgreSQL server: `47.117.131.22` (`db.refining.ltd`)
+- `secrets-mcp` app server: `47.238.146.244` (`secrets.refining.app`)
+
+## 1) Issue certificate for `db.refining.ltd` (Let's Encrypt + Cloudflare DNS-01)
+
+Install `acme.sh` on the PostgreSQL server and use a Cloudflare API token with DNS edit permission for the target zone.
+
+```bash
+curl https://get.acme.sh | sh -s email=ops@refining.ltd
+export CF_Token="your_cloudflare_dns_token"
+export CF_Zone_ID="your_zone_id"
+~/.acme.sh/acme.sh --issue --dns dns_cf -d db.refining.ltd --keylength ec-256
+```
+
+Install cert/key into a PostgreSQL-readable path:
+
+```bash
+sudo mkdir -p /etc/postgresql/tls
+sudo ~/.acme.sh/acme.sh --install-cert -d db.refining.ltd --ecc \
+  --fullchain-file /etc/postgresql/tls/fullchain.pem \
+  --key-file /etc/postgresql/tls/privkey.pem \
+  --reloadcmd "systemctl reload postgresql || systemctl restart postgresql"
+sudo chown -R postgres:postgres /etc/postgresql/tls
+sudo chmod 600 /etc/postgresql/tls/privkey.pem
+sudo chmod 644 /etc/postgresql/tls/fullchain.pem
+```
+
+## 2) Configure PostgreSQL TLS and access rules
+
+In `postgresql.conf`:
+
+```conf
+ssl = on
+ssl_cert_file = '/etc/postgresql/tls/fullchain.pem'
+ssl_key_file = '/etc/postgresql/tls/privkey.pem'
+```
+
+In `pg_hba.conf`, allow app traffic via TLS only (example):
+
+```conf
+hostssl  secrets-mcp  postgres  47.238.146.244/32  scram-sha-256
+```
+
+Keep a safe admin path (`local` socket or restricted source CIDR) before removing old plaintext `host` rules.
+
+Reload PostgreSQL:
+
+```bash
+sudo systemctl reload postgresql
+```
+
+## 3) Verify server-side TLS
+
+```bash
+openssl s_client -starttls postgres -connect db.refining.ltd:5432 -servername db.refining.ltd
+```
+
+The handshake should succeed and the certificate should match `db.refining.ltd`.
+
+## 4) Update `secrets-mcp` app server env
+
+Use environment values like:
+
+```bash
+SECRETS_DATABASE_URL=postgres://postgres:***@db.refining.ltd:5432/secrets-mcp
+SECRETS_DATABASE_SSL_MODE=verify-full
+SECRETS_ENV=production
+```
+
+If you use private CA instead of public CA, also set:
+
+```bash
+SECRETS_DATABASE_SSL_ROOT_CERT=/etc/secrets/pg-ca.crt
+```
+
+Restart `secrets-mcp` after updating env.
+
+## 5) Verify from app server
+
+Run positive and negative checks:
+
+- Positive: app starts, migrations pass, dashboard + MCP API work.
+- Negative:
+  - wrong hostname -> connection fails
+  - wrong CA file -> connection fails
+  - disable TLS on DB -> connection fails
+
+This ensures no silent downgrade to weak TLS in production.

scripts/migrate-v0.3.0.sql

@@ -0,0 +1,194 @@
+-- ============================================================================
+-- migrate-v0.3.0.sql
+-- Schema migration from v0.2.x → v0.3.0
+--
+-- Changes:
+--   • entries:         namespace → folder, kind → type; add notes column
+--   • audit_log:       namespace → folder, kind → type
+--   • entries_history: namespace → folder, kind → type; add user_id column
+--   • Unique index: (user_id, name) → (user_id, folder, name)
+--     Same name in different folders is now allowed; no rename needed.
+--
+-- Safe to run multiple times (fully idempotent).
+-- Preserves all data in users, entries, secrets.
+-- ============================================================================
+
+BEGIN;
+
+-- ── entries: rename namespace→folder, kind→type ──────────────────────────────
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'entries' AND column_name = 'namespace'
+    ) THEN
+        ALTER TABLE entries RENAME COLUMN namespace TO folder;
+    END IF;
+END $$;
+
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'entries' AND column_name = 'kind'
+    ) THEN
+        ALTER TABLE entries RENAME COLUMN kind TO type;
+    END IF;
+END $$;
+
+-- Set NOT NULL + default for folder/type in entries
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'entries' AND column_name = 'folder'
+    ) THEN
+        UPDATE entries SET folder = '' WHERE folder IS NULL;
+        ALTER TABLE entries ALTER COLUMN folder SET NOT NULL;
+        ALTER TABLE entries ALTER COLUMN folder SET DEFAULT '';
+    END IF;
+END $$;
+
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'entries' AND column_name = 'type'
+    ) THEN
+        UPDATE entries SET type = '' WHERE type IS NULL;
+        ALTER TABLE entries ALTER COLUMN type SET NOT NULL;
+        ALTER TABLE entries ALTER COLUMN type SET DEFAULT '';
+    END IF;
+END $$;
+
+-- Add notes column to entries if missing
+ALTER TABLE entries ADD COLUMN IF NOT EXISTS notes TEXT NOT NULL DEFAULT '';
+
+-- ── audit_log: rename namespace→folder, kind→type ────────────────────────────
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'audit_log' AND column_name = 'namespace'
+    ) THEN
+        ALTER TABLE audit_log RENAME COLUMN namespace TO folder;
+    END IF;
+END $$;
+
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'audit_log' AND column_name = 'kind'
+    ) THEN
+        ALTER TABLE audit_log RENAME COLUMN kind TO type;
+    END IF;
+END $$;
+
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'audit_log' AND column_name = 'folder'
+    ) THEN
+        UPDATE audit_log SET folder = '' WHERE folder IS NULL;
+        ALTER TABLE audit_log ALTER COLUMN folder SET NOT NULL;
+        ALTER TABLE audit_log ALTER COLUMN folder SET DEFAULT '';
+    END IF;
+END $$;
+
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'audit_log' AND column_name = 'type'
+    ) THEN
+        UPDATE audit_log SET type = '' WHERE type IS NULL;
+        ALTER TABLE audit_log ALTER COLUMN type SET NOT NULL;
+        ALTER TABLE audit_log ALTER COLUMN type SET DEFAULT '';
+    END IF;
+END $$;
+
+ALTER TABLE audit_log DROP COLUMN IF EXISTS actor;
+
+-- ── entries_history: rename namespace→folder, kind→type; add user_id ─────────
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'entries_history' AND column_name = 'namespace'
+    ) THEN
+        ALTER TABLE entries_history RENAME COLUMN namespace TO folder;
+    END IF;
+END $$;
+
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'entries_history' AND column_name = 'kind'
+    ) THEN
+        ALTER TABLE entries_history RENAME COLUMN kind TO type;
+    END IF;
+END $$;
+
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'entries_history' AND column_name = 'folder'
+    ) THEN
+        UPDATE entries_history SET folder = '' WHERE folder IS NULL;
+        ALTER TABLE entries_history ALTER COLUMN folder SET NOT NULL;
+        ALTER TABLE entries_history ALTER COLUMN folder SET DEFAULT '';
+    END IF;
+END $$;
+
+DO $$ BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'entries_history' AND column_name = 'type'
+    ) THEN
+        UPDATE entries_history SET type = '' WHERE type IS NULL;
+        ALTER TABLE entries_history ALTER COLUMN type SET NOT NULL;
+        ALTER TABLE entries_history ALTER COLUMN type SET DEFAULT '';
+    END IF;
+END $$;
+
+ALTER TABLE entries_history ADD COLUMN IF NOT EXISTS user_id UUID;
+ALTER TABLE entries_history DROP COLUMN IF EXISTS actor;
+
+-- ── secrets_history: drop actor column ───────────────────────────────────────
+ALTER TABLE secrets_history DROP COLUMN IF EXISTS actor;
+
+-- ── Rebuild unique indexes: (user_id, folder, name) ──────────────────────────
+-- Note: folder is now part of the key, so same name in different folders is
+-- naturally distinct — no rename of existing rows needed.
+DROP INDEX IF EXISTS idx_entries_unique_legacy;
+DROP INDEX IF EXISTS idx_entries_unique_user;
+
+CREATE UNIQUE INDEX IF NOT EXISTS idx_entries_unique_legacy
+    ON entries(folder, name)
+    WHERE user_id IS NULL;
+
+CREATE UNIQUE INDEX IF NOT EXISTS idx_entries_unique_user
+    ON entries(user_id, folder, name)
+    WHERE user_id IS NOT NULL;
+
+-- ── Replace old namespace/kind indexes with folder/type ──────────────────────
+DROP INDEX IF EXISTS idx_entries_namespace;
+DROP INDEX IF EXISTS idx_entries_kind;
+DROP INDEX IF EXISTS idx_audit_log_ns_kind;
+DROP INDEX IF EXISTS idx_entries_history_ns_kind_name;
+
+CREATE INDEX IF NOT EXISTS idx_entries_folder
+    ON entries(folder) WHERE folder <> '';
+CREATE INDEX IF NOT EXISTS idx_entries_type
+    ON entries(type) WHERE type <> '';
+CREATE INDEX IF NOT EXISTS idx_entries_user_id
+    ON entries(user_id) WHERE user_id IS NOT NULL;
+CREATE INDEX IF NOT EXISTS idx_audit_log_folder_type
+    ON audit_log(folder, type);
+CREATE INDEX IF NOT EXISTS idx_entries_history_folder_type_name
+    ON entries_history(folder, type, name, version DESC);
+CREATE INDEX IF NOT EXISTS idx_entries_history_user_id
+    ON entries_history(user_id) WHERE user_id IS NOT NULL;
+
+COMMIT;
+
+-- ── Verification queries (run these manually to confirm) ─────────────────────
+-- SELECT column_name, data_type FROM information_schema.columns
+--   WHERE table_name = 'entries' ORDER BY ordinal_position;
+-- SELECT indexname, indexdef FROM pg_indexes WHERE tablename = 'entries';
+-- SELECT COUNT(*) FROM entries;
+-- SELECT COUNT(*) FROM users;
+-- SELECT COUNT(*) FROM secrets;