dsh-plugin-dev-kb 1.0.9 → 1.1.0

package/kb/site/en/reference/subsystems/web-client.md

@@ -58,7 +58,7 @@ This pairing is not a second source of business truth. Host controllers decide d
 
 `ui-session` installs the `session` scope adapter and publishes `useSessions`, `useSession`, `sessionId`, and `useProjection`. Domain adapters add further standard sources without putting React hooks on the model objects.
 
-`ui-conversation` binds once to each `SessionBinding.eventSource`. Its event registry correlates standard events and Client-only `chunkrow/*` history events into stable business Contexts, and its view registry materializes target snapshots. Packed runs stay single inputs and Matches through replay; Chat Assistant, Trajectory Assistant, and Turn Tail are the built-in Definitions that interpret them. `ui-chat` and `ui-trajectory` register separate Definitions and builders: they may interpret the same event family, but they do not import or share each other's final display model. The shell selects a registered view and passes its snapshot through standard hooks and Slots. [Conversation](./conversation.md) defines Context identity, replay, Location data, target builders, and keyed renderers.
+`ui-conversation` binds once to each `SessionBinding.eventSource`. Its event registry correlates durable Session events and Client-only `assistant/live-chunk` updates into stable business Contexts, and its view registry materializes target snapshots. Chat Assistant, Trajectory Assistant, and Turn Tail interpret both live chunks and the compact streams embedded in durable settlements, so reconnect and paged history reproduce the same Assistant state without durable token rows. `ui-chat` and `ui-trajectory` register separate Definitions and builders: they may interpret the same event family, but they do not import or share each other's final display model. The shell selects a registered view and passes its snapshot through standard hooks and Slots. [Conversation](./conversation.md) defines Context identity, replay, Location data, target builders, and keyed renderers.
 
 `ui-slots` provides the typed registry and lifecycle ledger; `ui-renderer` is the only package that binds bare observables through `useSyncExternalStore`, owns React contexts, and renders the root tree. Feature components receive framework hooks, owner props, store actions, and explicit injection through their derived props. [Web Client Slots](./slots.md) lists those inputs, extension APIs, and the current Slot hierarchy.
 

package/kb/site/en/reference/subsystems/workspace.md

@@ -43,7 +43,7 @@ interface Workspace {
    */
   readonly path: string
 
-  /** Display title. Defaults to `basename(path)` at create; duplicates are allowed. */
+  /** Display title. Defaults to the final path segment, or a filesystem root's own spelling; duplicates are allowed. */
   readonly title: string
 
   /** ISO-8601 creation instant, stamped at create and never rewritten. */
@@ -120,7 +120,7 @@ Ownership truth is the record's ordered `sessionIds`, never derived from session
 
 ## The registry: `ctx.workspaceRegistry`
 
-`WorkspaceRegistry` ([signatures](#ctxworkspaceregistry--workspaceregistry)) owns registration and resolution. `create(path, title?)` canonicalizes the path, rejects a nonexistent path (the original `ENOENT`) or a non-directory, returns the existing entity unchanged when the canonical path is already owned, and otherwise creates a record with `title ?? basename(path)` prepended to the durable registry order (different canonical paths may share a display title). `get(id)` and the ordered `list()` are synchronous cache reads; `resolveByPath(path)` applies the same realpath canon without creating. `delete(id)` removes only the registration, order entry, and session account — the directory, user files, live sessions, and persisted logs are never touched, so those sessions become Ungrouped ([decision](https://github.com/deepseek-ai/deepseek-harness/blob/master/.agents/notes/implemented/feature/2026-07-27-workspace-registration-deletion.md)); unknown ids return `false`. Create and delete persist a pending-mutation marker before their two writes (record + order) can diverge; startup resolves exactly the marked mutation — by deleting the marked table row, which completes an interrupted delete and rolls back an interrupted create (the registration is re-creatable, so rollback is the safe direction) — and an unmarked order/table mismatch fails loud as corruption.
+`WorkspaceRegistry` ([signatures](#ctxworkspaceregistry--workspaceregistry)) owns registration and resolution. `create(path, title?)` requires a fully qualified path, canonicalizes it, rejects a nonexistent path (the original `ENOENT`) or a non-directory, returns the existing entity unchanged when the canonical path is already owned, and otherwise creates a record with `title ?? defaultWorkspaceTitle(path)` prepended to the durable registry order (different canonical paths may share a display title, and a path with no final segment uses its root spelling). `get(id)` and the ordered `list()` are synchronous cache reads; `resolveByPath(path)` applies the same fully qualified realpath canon without creating. `delete(id)` removes only the registration, order entry, and session account — the directory, user files, live sessions, and persisted logs are never touched, so those sessions become Ungrouped ([decision](https://github.com/deepseek-ai/deepseek-harness/blob/master/.agents/notes/implemented/feature/2026-07-27-workspace-registration-deletion.md)); unknown ids return `false`. Create and delete persist a pending-mutation marker before their two writes (record + order) can diverge; startup resolves exactly the marked mutation — by deleting the marked table row, which completes an interrupted delete and rolls back an interrupted create (the registration is re-creatable, so rollback is the safe direction) — and an unmarked order/table mismatch fails loud as corruption.
 
 Sessions get their cwd at create time from whoever creates them, not from this registry — the API gateway resolves a new session's cwd from the chosen workspace's `path` (falling back to an explicit or default cwd), creates the session so the cwd lands in its immutable [`SessionHeader`](./persistence.md#sessionheader--metadata-beside-the-log), then calls `attachSession`, which re-validates that stored header cwd against the workspace path. On the first successful start, the registry bootstraps history from persisted headers alone (`id`, `cwd`, `createdAt` — never event bodies), grouping sessions with a valid canonical cwd into per-directory workspaces, newest first; the initialized marker is written last so an interrupted bootstrap resumes safely. The bootstrap is one-time: cwd-less legacy sessions stay Ungrouped, and sessions created afterwards join a workspace only through `attachSession`.
 
@@ -253,13 +253,13 @@ Durable workspace registry. Startup waits for `sessionPersistence`, builds one c
 
 ```ts cordis-catalog
 /**
- * Create or reuse a workspace for an existing directory. The path is
- * canonicalized through `fs.realpath`; a nonexistent path rejects with the
- * original error and a non-directory rejects. Repeated calls for the same
- * canonical path return the existing entity without changing its title.
+ * Create or reuse a workspace for an existing directory. The fully qualified
+ * path is canonicalized through `fs.realpath`; a relative, nonexistent, or
+ * non-directory path rejects. Repeated calls for the same canonical path
+ * return the existing entity without changing its title.
  * A newly created workspace is prepended to the durable registry order.
  * Different canonical paths may share a display title.
- * @param path - Existing directory to own, in any path spelling.
+ * @param path - Existing directory to own, in a fully qualified path spelling.
  * @param title - Display title used only when a new record is created.
  * @returns the existing or newly durable workspace.
  */
@@ -312,7 +312,7 @@ archiveSession(sessionId: SessionId): Promise<void>
  * Resolve by canonical directory path without creating or mutating a
  * workspace. A missing path rejects during `realpath`; an existing unowned
  * directory returns `undefined`.
- * @param path - Existing directory path in any spelling.
+ * @param path - Existing directory path in a fully qualified spelling.
  * @returns the workspace owning the canonical path, when one exists.
  */
 async resolveByPath(path: string): Promise<Workspace | undefined>

package/kb/site/en/reference/tool-catalog.md

@@ -39,9 +39,8 @@ This table connects model-visible tool names to the plugin package and service s
 | `@deepseek-ai/dsh-tool-session-query` | `session_event_read`, `session_event_search`, `session_event_trace`, `session_search`, `session_trace` | `ctx.tools`, `ctx.systemPrompt`, `ctx.sessionQuery`, `a calling Agent for workspace authority` | `tool/call`, `tool/result` | - | The five read-only tools hide provider cursors and authorize every result from the immutable calling agent session. The package is opt-in; compositions that need enforced deadlines or bounded inline output also mount the generic timeout or spill policies. |
 | `@deepseek-ai/dsh-tool-subagent` | `list_subagent_models`, `subagent` | `ctx.tools`, `ctx.subagents`, `ctx.systemPrompt`, `ctx.llm for model discovery and selected-route validation` | `tool/call`, `tool/result`, `child session events through the chosen provider` | `subagent`, `subagent_fork` | The registered delegation name is the load-time `toolName` config (default `subagent`); the default schema above has model selection off, while the discovery schema is shown as the fixed companion available in an enabled Session. Web presets sample the Plugins preference for each new top-level Session and preserve that decision for its child Sessions; `subagent_fork` remains fixed-route. Each instance independently controls whether it reads model-selection settings and its background behavior through `modelSelectionSettings`, `backgroundMode`, and `enableRunInBackground`. |
 | `@deepseek-ai/dsh-tool-subagent-control` | `interrupt_agent`, `list_agents`, `send_message` | `ctx.tools`, `ctx.subagents`, `ctx.agents and ctx.sessionProjections (list_agents only)` | `tool/call`, `tool/result`, `child session events through ctx.subagents` | - | The globally named control tools over continuable background subagents: provider-bound `tool-subagent` instances register distinct delegation tools, while this package registers `send_message` and `interrupt_agent` once, plus `list_agents` from its separately loaded `/list-agents` plugin (whose catalog rows use the sessionProjections and live Agent registries). |
-| `@deepseek-ai/dsh-tool-subagent-report` | `report` | `ctx.subagents`, `ctx.systemPrompt`, `a live continuable in-process child Agent` | `tool/call`, `tool/result`, `a user-role message in the direct parent session` | - | Registered per continuable in-process child rather than globally, so this schema is visible only inside such a child and survives its global `toolFilter`. The same contribution installs the child-scoped `tool:report` prompt section, which this catalog does not render. The parent-facing `send_message` tool is installed independently. |
 | `@deepseek-ai/dsh-tool-jobs` | `job_kill`, `job_list`, `job_output` | `ctx.tools`, `ctx.jobs`, `ctx.systemPrompt` | `tool/call`, `tool/result`, `user/message via agent.inject() for background completion notices` | - | The kind-agnostic background-job controller: background bash commands, PTY sends, and subagents are read, listed, and killed through the same three tools. Loading the plugin attaches the controller that arms producers' `ctx.jobs.start()`. |
-| `@deepseek-ai/dsh-experimental-tool-agent-team` | `followup_task`, `interrupt_agent`, `list_agents`, `send_message`, `spawn_teammate`, `team_task_create`, `team_task_get`, `team_task_list`, `team_task_update`, `wait_agent` | `ctx.tools`, `ctx.systemPrompt`, `ctx.agentTeams`, `an exact live Team member Agent` | `tool/call`, `team/member`, `team/message/queued`, `team/message/delivered`, `team/task`, `tool/result` | - | All ten tools are scoped to implicit Team Leads and durable teammates. The shipped dsh-base bundle keeps the package disabled; the documented Agent Teams profile patch enables it while disabling the legacy continuable-child control names. |
+| `@deepseek-ai/dsh-experimental-tool-agent-team` | `interrupt_agent`, `list_agents`, `send_message`, `spawn_teammate`, `team_task_create`, `team_task_get`, `team_task_list`, `team_task_update`, `wait_agent` | `ctx.tools`, `ctx.systemPrompt`, `ctx.agentTeams`, `an exact live Team member Agent` | `tool/call`, `team/member`, `team/message/queued`, `team/message/delivered`, `team/task`, `tool/result` | - | All nine tools are scoped to implicit Team Leads and durable teammates. The shipped dsh-base bundle keeps the package disabled; the documented Agent Teams profile patch enables it while disabling the legacy continuable-child control names. |
 | `@deepseek-ai/dsh-tool-todo` | `todo_write` | `ctx.tools`, `owning Agent session` | `tool/call`, `todo/write`, `tool/result` | - | todo_write is session-owned state; UIs render the latest todo/write event as a checklist. `allowParallelInProgress` is required with no default, so the catalog states its choice: `true`, whose description invites several `in_progress` items. A deployment choosing `false` receives the same tool with a description asking for exactly one active task. |
 | `@deepseek-ai/dsh-tool-workflow` | `workflow` | `ctx.tools`, `ctx.workflowEngine`, `ctx.systemPrompt`, `a calling Agent (exec.agent parents the script children)` | `tool/call`, `tool/result` | - | - |
 | `@deepseek-ai/dsh-tool-web` | `web_fetch`, `web_search` | `ctx.tools`, `ctx.web`, `ctx.systemPrompt` | `tool/call`, `tool/result` | - | web_search and web_fetch keep provider selection behind ctx.web so model-visible schemas stay stable across backend swaps. |
@@ -735,7 +734,7 @@ Source: [`packages/fs/tool-fs/src/index.ts`](https://github.com/deepseek-ai/deep
 
 ### `read_image`
 
-Read a PNG/JPEG/WebP/GIF file and return the image itself. Harness validates and downscales large supported images before the next model request, so use this tool directly instead of installing image libraries or creating thumbnails merely to inspect an image. Independent files may be read concurrently in small batches. Requires the current model to accept image input.
+Read a PNG/JPEG/WebP/GIF file and return the image itself. A path without a file extension is accepted; the format is detected from the file content, so normalized attachment paths can be passed directly without copying or renaming. Harness validates and downscales large supported images before the next model request, so use this tool directly instead of installing image libraries or creating thumbnails merely to inspect an image. Independent files may be read concurrently in small batches. Requires the current model to accept image input.
 
 ```json
 {
@@ -1622,7 +1621,7 @@ Source: [`packages/subagent/tool-subagent-control/src/index.ts`](https://github.
 
 ### `list_agents`
 
-List your continuable background subagents by durable id and label. Use it to recall which ones you started, not to poll for completion — you are told when one finishes. Status comes from the live registry: running means the agent is working right now, idle means it is loaded but between turns (it may be waiting on agents it started), and ready means it exists only in storage — resumable, not terminal, and not a result waiting to be collected; a `send_message` starts a new turn on the same conversation, and a direct child remains a `send_message` candidate in every status. The snapshot is not a delivery promise — `send_message` performs the authoritative check and may still fail. Children that could not be read are reported as diagnostics instead of being silently dropped. Scope `descendants` walks the whole tree below you in stable pre-order, annotating each entry with its durable direct-parent session id and depth. You may use `send_message` only for depth-1 entries; deeper entries are candidates for `interrupt_agent` only.
+List your continuable background subagents by durable id and label. Use it to recall which ones you started, not to poll for completion — you are told when one finishes. Status comes from the live registry: running means the agent is working right now, idle means it is loaded but between turns (it may be waiting on agents it started), and ready means it exists only in storage — resumable, not terminal, and not a result waiting to be collected; a `send_message` steers a running child at its nearest step boundary or starts a turn for an idle or ready child, and a direct child remains a `send_message` candidate in every status. The snapshot is not a delivery promise — `send_message` performs the authoritative check and may still fail. Children that could not be read are reported as diagnostics instead of being silently dropped. Scope `descendants` walks the whole tree below you in stable pre-order, annotating each entry with its durable direct-parent session id and depth. You may use `send_message` only for depth-1 entries; deeper entries are candidates for `interrupt_agent` only.
 
 ```json
 {
@@ -1644,23 +1643,23 @@ Source: [`packages/subagent/tool-subagent-control/src/list-agents.ts`](https://g
 
 ### `send_message`
 
-Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.
+Send a message to a direct continuable child by its agent id. If you are a resident continuable child, you may also target your direct parent. If the target is still working, the message steers its nearest step; if it is idle, the message starts a turn. This call returns no answer from the agent — only confirmation that the message was delivered. A failure means the message was NOT delivered.
 
 ```json
 {
   "type": "object",
   "properties": {
-    "subagent_id": {
+    "agent_id": {
       "type": "string",
-      "description": "The subagent id returned when the background subagent was started."
+      "description": "The agent id of your direct continuable child, or your direct parent when you are a resident continuable child."
     },
     "message": {
       "type": "string",
-      "description": "The message to deliver to the subagent."
+      "description": "The message to deliver to the agent."
     }
   },
   "required": [
-    "subagent_id",
+    "agent_id",
     "message"
   ]
 }
@@ -1670,33 +1669,6 @@ Source: [`packages/subagent/tool-subagent-control/src/index.ts`](https://github.
 
 The globally named control tools over continuable background subagents: provider-bound `tool-subagent` instances register distinct delegation tools, while this package registers `send_message` and `interrupt_agent` once, plus `list_agents` from its separately loaded `/list-agents` plugin (whose catalog rows use the sessionProjections and live Agent registries).
 
-<a id="deepseek-aidsh-tool-subagent-report"></a>
-
-## `@deepseek-ai/dsh-tool-subagent-report`
-
-### `report`
-
-Report selected content to the agent that started you. Call this once before you finish, with a self-contained final result, and earlier for progress or findings that change what that agent does next. That agent shares your workspace but does not automatically receive your transcript, tool output, or reasoning, so finishing your work is not itself a result. Reporting does not end your turn or finish your work, and only your direct parent receives it. A failed call may still have arrived, so do not blindly repeat it.
-
-```json
-{
-  "type": "object",
-  "properties": {
-    "output": {
-      "type": "string",
-      "description": "Actionable content for your parent; summarize conclusions and reference relevant shared paths."
-    }
-  },
-  "required": [
-    "output"
-  ]
-}
-```
-
-Source: [`packages/subagent/tool-subagent-report/src/index.ts`](https://github.com/deepseek-ai/deepseek-harness/blob/master/packages/subagent/tool-subagent-report/src/index.ts)
-
-Registered per continuable in-process child rather than globally, so this schema is visible only inside such a child and survives its global `toolFilter`. The same contribution installs the child-scoped `tool:report` prompt section, which this catalog does not render. The parent-facing `send_message` tool is installed independently.
-
 <a id="deepseek-aidsh-tool-jobs"></a>
 
 ## `@deepseek-ai/dsh-tool-jobs`
@@ -1774,32 +1746,6 @@ The kind-agnostic background-job controller: background bash commands, PTY sends
 
 ## `@deepseek-ai/dsh-experimental-tool-agent-team`
 
-### `followup_task`
-
-Send a durable follow-up task to another Team member and start a turn when needed.
-
-```json
-{
-  "type": "object",
-  "properties": {
-    "target": {
-      "type": "string",
-      "description": "Team member name, or lead."
-    },
-    "message": {
-      "type": "string",
-      "description": "Self-contained message for the target."
-    }
-  },
-  "required": [
-    "target",
-    "message"
-  ]
-}
-```
-
-Source: [`packages/experimental/tool-agent-team/src/index.ts`](https://github.com/deepseek-ai/deepseek-harness/blob/master/packages/experimental/tool-agent-team/src/index.ts)
-
 ### `interrupt_agent`
 
 Interrupt one teammate's current turn while preserving its pending inbox. Team Lead only.
@@ -1836,7 +1782,7 @@ Source: [`packages/experimental/tool-agent-team/src/index.ts`](https://github.co
 
 ### `send_message`
 
-Send durable information to another Team member without starting an idle member.
+Send one durable message to another Team member. A running target receives it at the nearest step boundary; an idle target starts a turn; an inactive teammate cold-resumes.
 
 ```json
 {
@@ -2084,7 +2030,7 @@ Wait for the next teammate status, mailbox, or shared-task change after this cal
 
 Source: [`packages/experimental/tool-agent-team/src/index.ts`](https://github.com/deepseek-ai/deepseek-harness/blob/master/packages/experimental/tool-agent-team/src/index.ts)
 
-All ten tools are scoped to implicit Team Leads and durable teammates. The shipped dsh-base bundle keeps the package disabled; the documented Agent Teams profile patch enables it while disabling the legacy continuable-child control names.
+All nine tools are scoped to implicit Team Leads and durable teammates. The shipped dsh-base bundle keeps the package disabled; the documented Agent Teams profile patch enables it while disabling the legacy continuable-child control names.
 
 <a id="deepseek-aidsh-tool-todo"></a>
 

package/kb/site/guide/network-proxy.md

@@ -0,0 +1,87 @@
+---
+editSource: "docs/user/guide/network-proxy.zh.md"
+---
+
+# 在网络代理后面运行 DSH
+
+DSH 会把自身的出站请求——模型调用、web 搜索、页面抓取、走 HTTP 的 MCP 服务器——都经由标准代理环境变量所指定的代理发出。它在启动时读取这些变量，不需要其他配置。有几条路径出于设计或运行时限制保持直连，下文"哪些保持直连"一节列出了它们。
+
+## 导出环境变量
+
+```sh
+export HTTPS_PROXY=http://127.0.0.1:7890
+export HTTP_PROXY=http://127.0.0.1:7890
+```
+
+把这两行写进 shell 配置，这样每次调用 `dsh` 都会继承它们；也可以写进 `$DSH_HOME/.env`（默认 `~/.dsh/.env`），和 API key 放在一起；导出的环境变量始终优先于该文件。项目自己的 `.env` 不能设置它们：它随 `git clone` 一起到来，DSH 宁可拒绝启动，也不让一个仓库决定你的流量去向。
+
+需要凭据的代理把凭据写在 URL 里：`http://user:password@proxy.example:8080`。DSH 绝不会回显这个 URL：诊断只点名被拒绝的变量，因此用户名和密码都不会出现在任何地方。
+
+## 为什么浏览器走代理、终端却不走
+
+这是最常见的意外，而且并非 DSH 特有。**根本不存在一个所有软件都遵循的"系统代理"**——实际上有三套互不相干的机制：
+
+| 机制 | 谁会遵循 |
+|---|---|
+| 操作系统的代理设置 | Safari、绝大多数 macOS 原生应用、Chrome 与 Edge |
+| `HTTP_PROXY` / `HTTPS_PROXY` 环境变量 | `curl`、`git`、`npm`、`pip` 以及 DSH |
+| TUN 模式（虚拟网卡） | 所有程序，且对应用透明 |
+
+Clash 这类代理软件里的"系统代理"开关只写第一套。浏览器会读到它，命令行工具则永远看不到。这就是为什么导出环境变量是一个独立步骤，也是为什么打开 TUN 模式后两者都能工作、且完全不需要变量。
+
+DSH 不读取操作系统的代理设置。请导出环境变量，或使用 TUN 模式。
+
+## 指定哪些目标保持直连
+
+`NO_PROXY` 列出需要直连的主机：
+
+```sh
+export NO_PROXY=internal.example.com,.corp.example.com,registry.local
+```
+
+一个条目写的是主机名，它连同其下所有子域名一起匹配：`NO_PROXY=example.com` 也会让 `api.example.com` 直连。前缀 `.` 或 `*.` 可以写，含义相同。条目可带 `:port`，`*` 则放行全部。
+
+**CIDR 网段不生效。** 操作系统的绕过列表常含 `10.0.0.0/8` 或 `192.168.0.0/16` 这类条目；把它们复制进 `NO_PROXY` 不会有任何效果。请改用主机名或域名后缀。
+
+不需要列出 `localhost` 或 `127.0.0.1`。DSH 始终绕过 loopback，否则它自己的 Web UI 与本地服务器都会经由代理并形成回环。
+
+## 值得知道的限制
+
+**不支持 SOCKS 代理。** `socks5://` 形式的值会在启动时被报告并跳过，指定它的那个 scheme 转为直连——把 `HTTPS_PROXY=socks5://…` 与一个可用的 `HTTP_PROXY` 一起设置时，`https:` 会保持直连，而不会去借用 HTTP 代理。请把变量指向代理软件的 HTTP 端口——多数软件两者都提供，且 HTTP 端口通常就在相邻的端口号上。
+
+**只设 `ALL_PROXY` 也够用。** DSH 会用它为两种协议兜底，尽管 Node 与 curl 在这一点上并不一致。显式设置 `HTTPS_PROXY` 仍然更清楚。
+
+**做 TLS 拦截的企业代理需要它的证书。** 如果代理已经可达但请求仍报证书错误，请在启动前把 Node 指向你所在组织的 CA 包：
+
+```sh
+export NODE_EXTRA_CA_CERTS=/path/to/corporate-ca.pem
+```
+
+Node 只在进程启动时读取该变量，所以要在运行 `dsh` 之前导出。
+
+**DSH 替你运行的工具遵循同一个代理。** bash 工具里的命令、`git`、`gh`，以及作为子进程启动的 MCP 服务器都会继承这些变量。子进程若本身是 Node 程序，则需 Node 22.21 或更高版本才会遵循；更旧的 Node 会直连。如果你的某个代理变量是 DSH 拒绝的值——比如 SOCKS URL——基于 Node 的工具同样直连而不是起不来，`curl` 与 `git` 则仍会读取那个值。
+
+**代理 URL 里的密码同样会到达这些工具。** `HTTPS_PROXY=http://alice:s3cret@proxy.example:8080` 就是一个普通环境变量，因此 DSH 运行的每一条命令——包括模型编写的那些——都能读到它，而打印环境的命令会把密码写进被保留的输出。这与该变量在你 shell 里对其他一切程序的行为一致。若这一点重要，请为代理提供一个无需凭据的入口，或改用 URL 之外的方式认证。
+
+## 哪些保持直连
+
+并非 DSH 发出的每个请求都会走代理：
+
+- **本机上的一切。** loopback 始终直连：`localhost`、整个 `127.0.0.0/8` 段、`::1` 与 `0.0.0.0`。代理无法有意义地访问一个只在本地监听的服务。
+- **模型编写的代码。** workflow 与 code-runtime worker 从不接收代理配置，因此模型编写的脚本读不到可能携带密码的代理 URL。这类脚本只有自行配置才能联网。
+- **使用情况遥测。** OTLP 导出器用的是 Node 自带的 HTTP 客户端，而不是代理所配置的那个，因此遥测直连；在禁止直连出网的环境里它只会失败。DSH 的任何功能都不依赖它。设 `DSH_TELEMETRY_MODE=DISABLED` 可完全关闭。
+- **`web_fetch` 访问字面量私网地址。** 形如 `http://10.0.0.5/` 的 URL 会被拒绝而非交给代理，与未配置代理时得到的拒绝相同。
+
+## 验证是否生效
+
+让 agent 抓取一个页面，同时观察代理软件的连接日志：
+
+```sh
+dsh --profile headless "fetch https://example.com and tell me the page title"
+```
+
+如果请求没有出现在那里，确认变量确实进入了 DSH 自己的环境：
+
+```sh
+env | grep -i proxy
+```

package/kb/site/guide/providers.md

@@ -14,21 +14,39 @@ editSource: "docs/user/guide/providers.zh.md"
 
 密钥是只写的。保存后，页面只会收到脱敏描述符，永远不会收到明文密钥。密钥存储在 `$DSH_HOME/.credentials.yaml` 中，settings 只保留它的凭据引用。
 
-## 添加目录提供方
+## 添加内置提供方
 
-选择**添加提供方**，选取 Anthropic 或 OpenAI 等提供方，输入其 API 密钥并保存。已安装目录会提供端点、协议和模型列表。
+选择**添加提供方**，选取 dsh 自带的提供方；列表显示的是提供方 id，例如 `anthropic`、`openai`、Kimi 对应的 `moonshotai`、GLM 对应的 `zai`。输入其 API 密钥并保存。已安装目录会提供端点、协议和模型列表。
 
-使用原生认证的提供方需要各自的原生凭据。Bedrock、Vertex、Azure 和 Codex 分别使用 AWS 凭据与区域、ADC 项目、`api-version` 和 OAuth；只填写 API 密钥字段无法完成配置。
+通过 OAuth 登录的提供方（例如 Codex）暂不支持。
 
 ## 添加自定义提供方
 
-对于公司网关、自建服务器或已安装目录中不存在的提供方，选择**添加自定义提供方**。提供小写 Provider ID、基础 URL、API 协议、凭据和至少一个模型。
+对于公司网关、自建服务器或已安装目录中不存在的提供方，选择**添加自定义提供方**。提供小写 Provider ID、基础 URL、API 协议、凭据和至少一个模型。**API 协议**必须选网关实际使用的那一种，表单提供三种：`openai-completions` 对应 OpenAI Chat Completions，`openai-responses` 对应 OpenAI Responses API，`anthropic-messages` 对应 Anthropic Messages API。一个提供方只使用一种协议，网关同时提供两种时需要建两个提供方。
 
 ![自定义提供方表单：Provider ID、显示名称、API 地址、API 协议、API 密钥](./providers-custom-form.zh.png)
 
 Provider ID 是永久的，因为请求、已保存会话、模型默认值和凭据引用都会使用它。如需重命名提供方，请添加新提供方并删除旧提供方。显示名称、基础 URL、协议、凭据和模型仍可编辑。
 
-在**模型目录**中选择**获取可用模型**，可查询表单当前显示的基础 URL 和凭据。选择候选项只会更新草稿；保存前不会存储提供方。目录提供方使用已安装目录，不发起网络请求。
+### 探测模型
+
+在**模型目录**中选择**获取可用模型**，即可询问端点它提供哪些模型。请求使用表单当前的 API 地址、协议和密钥，已保存的提供方则用已存储的密钥；响应会打开一个可搜索的选择框，搜索、勾选想要的模型，再点**添加所选**。保存或创建提供方之前不会存储任何内容。
+
+探测读取的是常见网关公开的列表格式，但并非每个端点都用这些格式作答，所以它只是便利手段而非保证：探测失败或列表为空时，手动添加模型 ID 即可，效果完全一样。内置提供方一律由已安装目录作答，即使其 API 地址指向网关也是如此，要查看网关实际提供的模型，请通过自定义提供方探测。
+
+## 选择模型
+
+已配置的提供方会出现在模型选择器中。选择模型也会将其设为新会话的默认值。已发送过请求的会话会保留自身日志中记录的模型。
+
+如果已保存默认值指向已删除的提供方，输入框会显示**选择模型**，并在选择其他模型前阻止输入。
+
+## 进阶配置
+
+自动生成的[插件配置目录](../reference/config-catalog.md)列出每个插件的所有受支持字段与默认值；[`dsh-llm-pi-ai`](../reference/config-catalog.md#deepseek-aidsh-llm-pi-ai) 就是本页所配置的那个提供方段落。[`dsh-llm-pi-ai`](https://github.com/deepseek-ai/deepseek-harness/blob/master/packages/llm/llm-pi-ai/README.zh.md) 和 [`dsh-llm-deepseek`](https://github.com/deepseek-ai/deepseek-harness/blob/master/packages/llm/llm-deepseek/README.zh.md) 参考文档负责直接 `settings.yaml` 配置、目录解析、推理控制、凭据与适配器错误。
+
+::: tip 表单刻意保持精简
+模型页只开放让一条路由得以存在的字段：API 密钥、显示名称、API 地址、API 协议，以及每个模型的 ID、显示名称、上下文窗口和最大输出 token 数。其余所有字段——推理等级、图片输入、请求兼容性开关、请求头、超时、重试策略——都在 `$DSH_HOME/settings.yaml` 中设置，也就是模型页写入的同一份文档。可以直接编辑它；浏览器与服务器在同一台机器时，也可以点击设置页顶部的**打开配置文件**打开它。适配器会在下一次请求时重新读取，无需重启任何东西。下面各小节介绍多数网关会用到的字段。
+:::
 
 ### 图片输入
 
@@ -66,7 +84,7 @@ llm-pi-ai:
         - id: second-model
 ```
 
-`defaultInput` 是回退值而不是覆盖值，默认为 `[text]`：在目录提供方上，它只为目录未描述的模型作答，因此绝不会把目录中本就具备图片能力的模型的该能力去掉。要收窄这类模型，请用它自己的 `input`。目录提供方没有可供填写的 `models` 列表，因此写在 `modelOverrides` 下，以模型 id 为键：
+`defaultInput` 是回退值而不是覆盖值，默认为 `[text]`：在内置提供方上，它只为其目录未描述的模型作答，因此绝不会把目录中本就具备图片能力的模型的该能力去掉。要收窄这类模型，请用它自己的 `input`。内置提供方没有可供填写的 `models` 列表，因此写在 `modelOverrides` 下，以模型 id 为键：
 
 ```yaml
 llm-pi-ai:
@@ -81,6 +99,48 @@ llm-pi-ai:
 
 这两个字段都是对你端点的断言，而不是对它的检查。声明了端点并不提供的图片能力的模型不会在这里被拦下，改由提供方拒绝该请求。
 
+### 推理等级
+
+对于声明了推理等级的模型，模型选择器会提供**推理等级**菜单。内置提供方的模型从已安装目录继承其等级。手动录入的模型不声明任何等级，因此模型菜单里不会出现推理等级项，由端点自身的默认值决定模型是否思考。请在 `$DSH_HOME/settings.yaml` 中用 `reasoningEfforts` 声明等级：
+
+```yaml
+llm-pi-ai:
+  providers:
+    my-gateway:
+      apiKeyEnv: GATEWAY_API_KEY
+      api: openai-completions
+      baseURL: https://gateway.example/v1
+      reasoning: high
+      models:
+        - id: my-reasoner
+          reasoningEfforts:
+            off:
+            high: high
+            max: max
+```
+
+每个键都是菜单提供的一个等级，其值是在协议上以 `reasoning_effort` 发送的写法，因此 `max: xhigh` 可以为自有一套词汇的网关重命名某个等级。只有 `off` 可以留空，因为对多数端点来说，不思考就是不传该参数。路由的 `reasoning` 是会话尚未选择等级时采用的等级；在选择器中选定某个等级后，它会与模型一起保存为新会话的默认值。
+
+留空的 `off` 什么都不发送，这只能让「按请求才思考」的模型停下来；给 `off` 一个值，则会把该值作为 `reasoning_effort` 发送。对于「不明确关闭就会思考」的模型——例如 OpenAI 兼容网关后面的 DeepSeek V4——需要 `compat.thinkingFormat: deepseek`：它让 `off` 发送 `thinking: {type: disabled}`，其他每个等级则在 effort 之外再发送 `thinking: {type: enabled}`：
+
+```yaml
+      models:
+        - id: deepseek-v4-pro
+          compat:
+            thinkingFormat: deepseek
+          reasoningEfforts:
+            off:
+            high: high
+            max: max
+```
+
+网关并不提供推理能力的内置提供方模型，可在 `modelOverrides` 下用 `reasoningEfforts: false` 去掉其等级；之后再为它选择等级会被拒绝并报 `UNSUPPORTED_REASONING_EFFORT`。DeepSeek 自身的路由不需要以上任何配置：其模型已经提供 `off`、`low`、`high` 和 `max`，`llm-deepseek.reasoningEffort` 设置选择器的起始默认值：
+
+```yaml
+llm-deepseek:
+  reasoningEffort: max
+```
+
 ### 请求兼容性
 
 网关可能持有可用的密钥、地址也通得到，却仍然拒绝每一个请求。pi-ai 依据端点的 URL 决定请求的形状——系统提示词由哪个角色承载、输出上限写在哪个字段、思考级别如何传输——而对于它无法识别的地址，会当作 OpenAI 本身来对待。多数 OpenAI 兼容网关至少会拒绝 OpenAI 所接受的某一样东西。
@@ -117,23 +177,16 @@ llm-pi-ai:
 
 全部开关、各自接受的取值，以及接受它们的协议，都列在[生成的 `dsh-llm-pi-ai` 配置参考](../reference/config-catalog.md#deepseek-aidsh-llm-pi-ai)的 `PiAiCompatProfile` 之下——该参考派生自源码，因此不会落后于适配器实际接受的内容。
 
-## 选择模型
-
-已配置的提供方会出现在模型选择器中。选择模型也会将其设为新会话的默认值。已发送过请求的会话会保留自身日志中记录的模型。
-
-如果已保存默认值指向已删除的提供方，输入框会显示**选择模型**，并在选择其他模型前阻止输入。
-
 ## 排错
 
 - **`MISSING_CREDENTIAL`**：通过模型页存储提供方密钥，或提供被引用的环境变量。
 - **`UNKNOWN_MODEL`**：选择已配置的模型，或向自定义提供方添加缺失的模型。
 - **获取可用模型返回 401**：检查密钥。模型发现会调用 OpenAI 兼容的 `GET /models` 端点；对于不提供该端点的服务，请手动输入模型。
+- **获取可用模型提示既没有 `data` 数组也没有 `models` 对象**：端点返回的列表格式不在探测的读取范围内。请手动输入模型。
 - **密钥与地址都正确，网关却拒绝每一个请求**：它的请求形状与 OpenAI 不同。先在路由上设 `compat.supportsDeveloperRole: false` 与 `compat.maxTokensField: max_tokens`。
 - **只有推理模型失败**：pi-ai 把它们的系统提示词以 `developer` 角色发出，而网关拒绝该角色。设 `compat.supportsDeveloperRole: false`。
+- **手动录入的模型没有推理等级菜单**：该模型没有声明任何等级。在 `settings.yaml` 中给该模型加上 `reasoningEfforts`。
+- **`off` 无法让 DeepSeek 模型停止思考**：留空的 `off` 不发送任何推理字段，默认思考的端点就继续思考。请在模型或路由上设置 `compat.thinkingFormat: deepseek`。
 - **某个 compat 开关因没有值而被拒绝**：冒号后什么都没写。给它一个值，或删掉该键以沿用已安装 catalog 的值。
-- **图片在发送前被拒绝**：该模型未声明图片模态。请给自定义提供方的模型加上 `input: [text, image]`；DeepSeek 自身的 chat-completions 路由是纯文本的，且无法通过配置改变。
+- **图片在发送前被拒绝**：该模型未声明图片模态。请给自定义提供方的模型加上 `input: [text, image]`；在 DeepSeek 自身的路由上，请选择声明了图片能力的模型 `deepseek-v4-flash-vision-exp`。
 - **提供方拒绝了带图片的请求**：该模型声明了其端点实际并不提供的图片能力。请从授予它图片能力的那个列表中移除 `image`——可能是模型的 `input`，也可能是路由的 `defaultInput`——然后开启新会话：附加的图片会留在会话日志里，因此在会话离开它之前，同一个请求会不断重复。
-
-## 进阶配置
-
-自动生成的[插件配置目录](../reference/config-catalog.md)列出每个插件的所有受支持字段与默认值；[`dsh-llm-pi-ai`](../reference/config-catalog.md#deepseek-aidsh-llm-pi-ai) 就是本页所配置的那个提供方段落。[`dsh-llm-pi-ai`](https://github.com/deepseek-ai/deepseek-harness/blob/master/packages/llm/llm-pi-ai/README.zh.md) 和 [`dsh-llm-deepseek`](https://github.com/deepseek-ai/deepseek-harness/blob/master/packages/llm/llm-deepseek/README.zh.md) 参考文档负责直接 `settings.yaml` 配置、目录解析、推理控制、凭据与适配器错误。

package/kb/site/guide/schedule.md

@@ -12,6 +12,8 @@ dsh web --patch apps/cli/config/examples/schedule/cordis.yml
 
 当前 overlay 支持使用正整数 `after_seconds`、绝对时间 `at` 目标，或至少 300 秒的固定速率 `every_seconds` 间隔创建提醒。模型通过 `schedule_create`、`schedule_list` 和 `schedule_delete` 管理它们；每个结果都会把交付标为 `session-local`。
 
+启用此 overlay 后，成功打开且存在活动提醒的 Session 会在对话 header 中显示只读目录。目录列出完整 prompt、等待中或已逾期状态、单次或精确重复周期、浏览器本地目标时间与相对时间。侧边栏还会在 grouped、flat 与 search 行当前可用的 projection 值非空时，于标题后显示不可交互的闹钟。这些界面不会创建、编辑、删除或确认提醒；cold Session 的缓存闹钟允许短暂漏显或残留。
+
 浏览器会为每条提示词附加其 IANA 时区。Time-context 会告诉模型，把未明确限定时区的日期和时间解释为该请求的浏览器时区。此假设仅用于自然语言解释：`schedule_create.at` 必须是带 `Z` 或数值偏移量且严格符合 RFC 3339 的日期时间，或是带显式 `UTC` 或 IANA Area/Location 时区的 `{ date, time, time_zone }`。Schedule 不保留或推断 Session 默认时区。夏令时缺口会被拒绝，重叠时段选择第一个时刻；成功创建的记录只保留所得的 UTC 目标。
 
 每条提醒由原 Session 日志拥有。live 根 Agent 会等待到完全 idle，再在该对话中排入一个普通 follow-up 轮次。它绝不会中途引导当前工作，也不会添加独立回执或提醒卡片。关闭进程或让 Session 保持 cold 会停止内存 timer，但不会删除记录；重新打开同一个 Session 会恢复等待并交付逾期提醒。查看 cold 历史不会激活提醒，fork 也不会继承父 Session 的提醒。

package/kb/site/reference/agent-lifecycle.md

@@ -39,14 +39,16 @@ sequenceDiagram
   Driver->>Prompt: <code>system-prompt/assemble</code> waterfall
   Driver->>LLM: <code>agent/request</code> waterfall, then <code>llm/stream</code> waterfall
   LLM-->>Driver: StreamChunk*
-  Driver->>Session: <code>assistant/chunk</code>*
-  Session-->>SDK: <code>session/event</code> <code>assistant/chunk</code>*
+  Driver-->>SDK: <code>agent/assistant-stream</code> chunk*
   alt final adapter or terminal in-band request failure
+    Driver->>Session: <code>assistant/attempt</code>
+    Driver-->>SDK: <code>agent/assistant-stream</code> committed end
     Driver->>Session: <code>step/end</code>
     Driver->>Hooks: <code>agent/request-error</code> waterfall
     Hooks-->>Driver: return retry action or preserve the original error
   else model request succeeded
   Driver->>Session: <code>assistant/message</code>
+  Driver-->>SDK: <code>agent/assistant-stream</code> committed end
   Driver->>Tools: classify pending call by executionMode
   loop barriers and bounded rolling pool, reclassify before start
     opt call starts
@@ -75,7 +77,7 @@ sequenceDiagram
   Driver-->>SDK: <code>agent/status</code> idle
 ```
 
-`assistant/message` 事件会记录每次成功的提供方调用，包括返回空内容或以 `max-tokens` 结束的调用。空内容不会进入派生历史，但该持久事件仍会保留用量，并通过 `sourceEventSeqs` 精确列出对应的 `assistant/chunk` 事件，包括显式空列表。
+`assistant/message` 事件会记录每次成功的提供方调用，包括返回空内容或以 `max-tokens` 结束的调用，并嵌入精确的紧凑带时间 stream。空内容不会进入派生历史。失败、重试、取消或 stream error attempt 到达 settlement 时，如果没有 surface message，就会把 stream 记录为 `assistant/attempt`。实时 `agent/assistant-stream` chunk frame 是瞬态数据；回放读取任一种持久 settlement，如果进程在 settlement 前硬中断，则不会留下持久 attempt stream。
 
 `dsh-compaction-basic` 在派生请求之前通过 `agent/pre-step` 处理压力，而 `agent/request-error` 仅用于规范的上下文溢出。任一触发条件满足后，系统都会先执行可选的工具结果剪枝，再选择摘要。恢复发生在失败步骤结束之后、失败轮次结束之前；只有当剪枝或摘要生成推进了 surface replacement generation 时，系统才会开启一个全新的重试轮次，否则仍以原始请求错误为准。
 

package/kb/site/reference/api-gateway.md

@@ -126,13 +126,13 @@ Connection 在 HTTP bridge 之前执行 `/api` 的统一信任检查，再在共
 
 Gateway 每次调用都从当前注册表解析描述符和实时服务，不缓存业务对象。它要求 `args` 的字段集合与描述符完全一致，先用 codec 校验 wire 值，再通过注册的 lookup 或 Context 提供方解析对象或接收者，最后调用 binding 指向的服务方法并校验返回值。缺少提供方、identity 未命中、binding 不一致、参数缺失或多余、schema 失败和方法不存在都会在进入业务代码前或离开业务代码后失败。
 
-lookup 提供方的 `register()` 同时提供稳定声明和默认 resolver；`configure()` 提供由 Host 组合拥有、可异步执行且受 effect 生命周期约束的 resolver。配置可以先于提供方挂载；没有提供方时调用仍以 `lookup-unavailable` 失败，配置卸载后则恢复提供方默认策略。Session Controller 负责 `agent` 与 `session` 的标准 `agentFor()` 语义：复用 live Agent，自动恢复普通冷会话，对并发恢复去重，并拒绝由 subagent routing 拥有的 identity；`session` lookup 返回该 Agent 的 Session。恢复失败和 ownership fence 通过既有 RPC error 原样返回，不折叠为 Gateway 的 `internal` 错误。
+lookup 提供方的 `register()` 同时提供稳定声明和默认 resolver；`configure()` 提供由 Host 组合拥有、可异步执行且受 effect 生命周期约束的 resolver。配置可以先于提供方挂载；没有提供方时调用仍以 `gateway/lookup-unavailable` 失败，配置卸载后则恢复提供方默认策略。Session Controller 负责 `agent` 与 `session` 的标准 resolver 语义：复用 live Agent，自动恢复普通冷会话，对并发恢复去重，并拒绝由 subagent routing 拥有的 identity；`session` lookup 返回该 Agent 的 Session。恢复失败与 ownership fence 抛出携带自有码的 `RemoteError`（`session/not-found` 或 `session/agent-busy`），Gateway 原样编码上 wire；只有未归类的 throw 才折成 `gateway/internal`。
 
 Client 卸载一个贡献时会一起移除描述符和具体方法，中止其进行中的调用，并使外部仍持有的陈旧方法句柄拒绝继续调用。Host 上已经注册过的严格 endpoint 被撤回后也不会降级到 SRC 推断，以免热卸载悄然降低校验强度。
 
 ## SRC 开发回退
 
-Host 通过 `node --import tsx/esm` 从源码启动时不会执行 Typert 编译插件。标准 decorator 初始化器仍会把方法名和调用模式记录到模块私有 `WeakMap`，`TypertRemoteService` 或 `bindTypertRemote()` 则提供显式服务 binding；Gateway 因而可以在不启动 `ts.Program` 的情况下构造一个较弱的临时描述符。
+Host 通过 `node --import tsx/esm` 从源码启动时不会执行 Typert 编译插件。标准 decorator 初始化器仍会把方法名和调用模式记录到 Service 原型上的带版本描述符中，`TypertRemoteService` 或 `bindTypertRemote()` 则提供显式服务 binding；Gateway 因而可以在不启动 `ts.Program` 的情况下构造一个较弱的临时描述符。描述符使用稳定的字符串属性名，因此 `remoteMethods()` 能读取协议包另一个已安装副本写入的标记。
 
 SRC 回退从运行中函数解析简单参数名。参数名与某个已注册 lookup 的 `parameter` 相同，例如 `agent` 或 `session`，就使用其 `agentId` 或 `sessionId` wire 字段并在 Host 解析对象；其他参数只检查值是否为无循环、无特殊 prototype 的 JSON-safe 数据。`@RemoteScope` 直接使用已注册 Host Context 提供方的 wire 字段。SRC 不读取 TypeScript 类型，不生成 Zod schema，不推断可选参数，也不支持解构、默认值、rest 或重复参数名。