Merge latest master into skill catalog hot refresh

2026-07-28 01:10:33 +08:00
parent 8c5d6468d9 dde4b3d518
commit 2612a4b587
1109 changed files with 33736 additions and 20360 deletions
--- a/packages/examples/acp-demo/README.i18n.yaml
+++ b/packages/examples/acp-demo/README.i18n.yaml
@@ -1,6 +1,6 @@
 # Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
-#   pnpm run verify-translation-pairing --write
-README.md: bbc41f1e0aa0c98a6e70ee54357675f1d7f05dbc
-README.zh.md: 03e1246d5358138c633d2b19a9c186a3beec5a1e
+#   pnpm run verify-translation-pairing --write packages/examples/acp-demo/README.md
+README.md: 395ab230146568989c4e6d1361218efb72d857e7
+README.zh.md: 1dfd2d99f4ab80953df77feba19649b934775d68
--- a/packages/examples/acp-demo/README.md
+++ b/packages/examples/acp-demo/README.md
@@ -36,7 +36,6 @@ The app does not install commands, user interaction, session navigation, configu
 | `toolBash` | owner defaults | Model-facing bash tool config. |
 | `toolTasks` | owner defaults | Generic background-task control config, or `false`. |
 | `goals` | owner defaults | Persisted same-session goal domain and model tools, or `false`. |
-| `llmRetry` | owner defaults | Bounded transient model-request retry policy. |

 The shipped [`examples/acp-agent/cordis.yml`](../../../examples/acp-agent/cordis.yml) adds the DeepSeek adapter, sandboxed bash and filesystem providers, one-shot approval policy, compaction, subagents, workflows, hooks, and model-facing tools. The app supplies the derived session-query index, while the model-facing query consumer remains an explicit leaf opt-in. Snapshot overlays replace only nondeterministic providers or policy values.

--- a/packages/examples/acp-demo/README.zh.md
+++ b/packages/examples/acp-demo/README.zh.md
@@ -36,7 +36,6 @@ ACP 自动化服务器应用：默认 agent 主干、客户端通过 [`@deepseek
 | `toolBash` | 拥有者默认值 | 面向模型的 bash 工具配置。 |
 | `toolTasks` | 拥有者默认值 | 通用后台任务控制配置，或 `false`。 |
 | `goals` | 拥有者默认值 | 持久的同会话目标领域与模型工具，或 `false`。 |
-| `llmRetry` | 拥有者默认值 | 有界的瞬时模型请求重试策略。 |

 已交付的 [`examples/acp-agent/cordis.yml`](../../../examples/acp-agent/cordis.yml) 添加 DeepSeek 适配器、沙箱化 bash 与文件系统提供方、一次性批准策略、压缩、subagent、工作流、钩子，以及面向模型的工具。应用提供派生会话查询索引，而面向模型的查询消费方仍由叶节点显式选用。快照 overlay 只替换非确定性提供方或策略值。

--- a/packages/examples/acp-demo/src/index.ts
+++ b/packages/examples/acp-demo/src/index.ts
@@ -69,8 +69,6 @@ export interface Config {
  toolTasks?: NonNullable<agentCore.Config['toolTasks']>
  /** Persisted same-session goals; owner defaults enable them, or false disables the stack and tools. */
  goals?: agentCore.GoalConfig | false
-  /** Bounded transient model-request retry policy forwarded through agent-core. */
-  llmRetry?: NonNullable<agentCore.Config['llmRetry']>
 }

 // Each front door owns a complete, directly readable config schema; extracting
@@ -96,7 +94,6 @@ export const Config: z<Config> = z.object({
  toolBash: agentCore.ToolBashConfigSchema,
  toolTasks: z.union([z.const(false), agentCore.ToolTasksConfigSchema]),
  goals: z.union([z.const(false), agentCore.GoalConfigSchema]),
-  llmRetry: agentCore.LlmRetryConfigSchema,
 })
 /* jscpd:ignore-end */

--- a/packages/examples/acp-demo/tests/acp-agent.spec.ts
+++ b/packages/examples/acp-demo/tests/acp-agent.spec.ts
@@ -1,12 +1,14 @@
 import { describe, expect, it } from 'vitest'
+import { randomUUID } from 'node:crypto'
 import { mkdtemp } from 'node:fs/promises'
 import { join } from 'node:path'
 import { tmpdir } from 'node:os'
 import { Context } from 'cordis'
 import Loader from '@cordisjs/plugin-loader'
-import { agentEvents, type Agent } from '@deepseek-ai/dsh-agent'
+import { agentEvents } from '@deepseek-ai/dsh-agent'
 import { TOOL_ORDER_REST } from '@deepseek-ai/dsh-system-prompt'
 import type { Message } from '@deepseek-ai/dsh-llm'
+import { SessionId } from '@deepseek-ai/dsh-session'
 import * as acpAgent from '../src/index.ts'

 /**
@@ -29,6 +31,7 @@ async function mount(config: acpAgent.Config, withBash = false): Promise<Context
      start() { throw new Error('composition test does not execute bash') },
    })
  }
+  config.persistenceRoot ??= await mkdtemp(join(tmpdir(), 'dsh-acp-demo-persistence-'))
  await ctx.plugin(acpAgent, config)
  return ctx
 }
@@ -42,12 +45,9 @@ async function isolatedSkillsConfig(catalogDescriptionMaxLength?: number): Promi
 }

 async function composePrefix(ctx: Context): Promise<Message[]> {
-  const agent = { session: { header: { cwd: '/tmp' } } } as unknown as Agent
-  const empty: Message[] = []
-  return await agentEvents(ctx, agent).waterfall(
-    'agent/session-prefix', empty, new AbortController().signal,
-    () => Promise.resolve(empty),
-  )
+  const agent = ctx.agentLoop.create(SessionId(`acp-demo-prefix-${randomUUID()}`), {}, { cwd: '/tmp' })
+  await agentEvents(ctx, agent).serial('agent/step', 1, 1, new AbortController().signal)
+  return agent.session.deriveMessages()
 }

 async function withIsolatedSkillHomes<T>(run: () => Promise<T>): Promise<T> {
--- a/packages/examples/acp-demo/tests/built-bin.e2e.ts
+++ b/packages/examples/acp-demo/tests/built-bin.e2e.ts
@@ -17,6 +17,7 @@ import {
 import { Readable, Writable } from 'node:stream'
 import { promisify } from 'node:util'
 import { zstdDecompress } from 'node:zlib'
+import { execa } from 'execa'
 import { afterEach, describe, expect, it } from 'vitest'

 /**
@@ -32,7 +33,7 @@ const decompress = promisify(zstdDecompress)
 const dshPackages = [
  'examples/agent-spine-demo', 'core/agent', 'core/session', 'core/system-prompt',
  'core/tools', 'core/agent-loop', 'llm/llm', 'bash/bash',
-  'bash/bash-local', 'bash/tool-bash', 'context/workspace-context', 'support/invariants', 'ui/app-boot',
+  'bash/bash-local', 'bash/tool-bash', 'subprocess/subprocess', 'subprocess/subprocess-local', 'context/workspace-context', 'support/invariants', 'ui/app-boot',
  'session-persistence/session-persistence',
  'session-persistence/session-checkpoint-policy', 'session-persistence/session-persistence-jsonl',
  'acp/acp', 'examples/acp-demo', 'util/paths',
@@ -94,6 +95,8 @@ async function makeConsumer(): Promise<string> {
  await writeFile(join(dir, 'cordis.yml'), [
    '- id: mock-llm',
    '  name: \'./mock-llm.mjs\'',
+    '- id: subprocess',
+    '  name: \'@deepseek-ai/dsh-subprocess-local\'',
    '- id: bash',
    '  name: \'@deepseek-ai/dsh-bash-local\'',
    '- id: acp-agent',
@@ -207,27 +210,22 @@ describe.skipIf(!existsSync(acpBin))('dsh-acp-demo BUILT bin (node lib/bin.js, n
    expect(code).not.toBe(0)
    expect(stderr).toContain('config file not found')
  }, 30_000)
+
 })

-/** Spawn the built acp bin against `configArg` and resolve with its exit code + stderr. */
-function runBinExpectingExit(configArg: string, cwd: string = tmpdir()): Promise<{ code: number; stderr: string }> {
-  return new Promise((resolve, reject) => {
-    const proc = spawn(process.execPath, [acpBin, '--config', configArg], {
-      cwd,
-      env: {
-        ...process.env,
-        DSH_HOME: join(cwd, '.dsh'),
-        DSH_AGENTS_HOME: join(cwd, '.agents'),
-      },
-      stdio: ['pipe', 'pipe', 'pipe'],
-    })
-    child = proc
-    let stderr = ''
-    proc.stderr.setEncoding('utf8')
-    proc.stderr.on('data', (c: string) => { stderr += c })
-    const timer = setTimeout(() => { proc.kill('SIGKILL'); reject(new Error(`bin did not exit within 25s. stderr:\n${stderr}`)) }, 25_000)
-    proc.on('exit', (code) => { clearTimeout(timer); resolve({ code: code ?? -1, stderr }) })
-    proc.on('error', (err) => { clearTimeout(timer); reject(err) })
-    proc.stdin.end()
+/** Spawn the built acp bin against `configArg` (stdin closed at EOF) and resolve with its exit code + stderr. */
+async function runBinExpectingExit(configArg: string, cwd: string = tmpdir()): Promise<{ code: number; stderr: string }> {
+  const result = await execa(process.execPath, [acpBin, '--config', configArg], {
+    cwd,
+    env: {
+      DSH_HOME: join(cwd, '.dsh'),
+      DSH_AGENTS_HOME: join(cwd, '.agents'),
+    },
+    input: '',
+    timeout: 25_000,
+    killSignal: 'SIGKILL',
+    reject: false,
  })
+  if (result.timedOut) throw new Error(`bin did not exit within 25s. stderr:\n${result.stderr}`)
+  return { code: result.exitCode ?? -1, stderr: result.stderr }
 }
--- a/packages/examples/acp-demo/tests/load-path.e2e.ts
+++ b/packages/examples/acp-demo/tests/load-path.e2e.ts
@@ -35,6 +35,8 @@ const CORDIS_YML = `
  name: '@deepseek-ai/dsh-llm-deepseek'
  config:
    apiKey: !!js process.env.DEEPSEEK_API_KEY
+- id: subprocess
+  name: '@deepseek-ai/dsh-subprocess-local'
 - id: bash
  name: '@deepseek-ai/dsh-bash-local'
 - id: acp-agent
--- a/packages/examples/agent-spine-demo/README.i18n.yaml
+++ b/packages/examples/agent-spine-demo/README.i18n.yaml
@@ -1,6 +1,6 @@
 # Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
-#   pnpm run verify-translation-pairing --write
-README.md: 32874bf2839c194572ddde8c4ed007297f763ccc
-README.zh.md: 57a06a00203b8e67f2f33c87d7450d1a0789d7e6
+#   pnpm run verify-translation-pairing --write packages/examples/agent-spine-demo/README.md
+README.md: 359e7153be2f480ba3fea4b06782acdc9f89ebb9
+README.zh.md: 57fec3f32f5bbc8f3d82ff8971d36d376d722753
--- a/packages/examples/agent-spine-demo/README.md
+++ b/packages/examples/agent-spine-demo/README.md
@@ -23,7 +23,7 @@ Read this package for the whole plugin tree and its composition order.
@deepseek-ai/dsh-goal             optional persisted same-session goal domain
@deepseek-ai/dsh-tool-goal        optional model-facing goal controls
@deepseek-ai/dsh-goal-session     optional same-session goal-round driver
-@deepseek-ai/dsh-llm-retry        bounded transient request retry policy
+@deepseek-ai/dsh-llm-retry        provider-routed request retry policy
@deepseek-ai/dsh-tasks-local      generic background-task registry
@deepseek-ai/dsh-invariants       configurable invariant registry service
@deepseek-ai/dsh-session/invariant
@@ -55,11 +55,11 @@ This is the [interface/implementation/consumer seam](../../../.agents/notes/impl

 ```ts
 import type { Config } from '@deepseek-ai/dsh-agent-spine-demo'
-// { agents?, maxParallelToolCalls?, persona?, toolOrder?, tools?, dshHome?, sessionTitle?, skills?, workspaceContext, toolBash?, toolTasks?, goals?, invariants?, llmRetry? }
+// { agents?, maxParallelToolCalls?, persona?, toolOrder?, tools?, dshHome?, sessionTitle?, skills?, workspaceContext, toolBash?, toolTasks?, goals?, invariants? }
 // workspaceContext requires { maxBytes } or false; the other owner schemas supply defaults.
 ```

-The bundle FORWARDS each field to the child that owns it: `agents` and `maxParallelToolCalls` to `agent-loop` (`agents` defaults to `[]`; the cap defaults there), so each app supplies its own pre-created agents — TUI and headless apps pre-create `main`, while the ACP app creates agents on demand at `session/new`; `llmRetry` to the bounded retry policy; `persona` and `toolOrder` to `dsh-system-prompt`; `tools` to the tool registry for its presentation mode; `sessionTitle` to the fallback title service; `skills.registry`, `skills.local`, and `skills.tool` to the skill registry, local provider, and model-facing consumer; the required `workspaceContext` choice to `dsh-workspace-context` (`{ maxBytes }` enables loading and `false` disables it); `invariants` to the invariant service; and `toolBash`/`toolTasks` to the two model-facing tool plugins the bundle owns. Omitted `sessionTitle` uses the explicit example policy of 5 words, 40 fallback bytes, and 80 accepted-title bytes. A `goals` object opts into the persisted domain, model tools, and same-session driver while forwarding `goals.domain` and `goals.tool` to their owners; omission or `false` leaves the stack absent so headless callers retain one-turn settlement. Set `skills.enabled: false` to omit both the local provider and model-facing skill tool, and set `toolTasks: false` to retain the task service for foreground producers without exposing `task_output` / `task_list` / `task_kill`. It resolves `dshHome` once through [`@deepseek-ai/dsh-paths`](../../util/paths/README.md) and forwards that absolute value to tool-bash's managed environment and enabled local skill discovery. An absent top-level `dshHome` adopts `skills.local.dshHome`; supplying both with different resolved paths fails loudly. `toolBash.enableRunInBackground` controls only the bash producer; independently loaded producers keep their own config. Workspace instructions register before the skill catalog so their session-prefix message renders first. App packages use `pickSpineConfig()` to copy only these bundle-owned fields.
+The bundle FORWARDS each field to the child that owns it: `agents` and `maxParallelToolCalls` to `agent-loop` (`agents` defaults to `[]`; the cap defaults there), so each app supplies its own pre-created agents — TUI and headless apps pre-create `main`, while the ACP app creates agents on demand at `session/new`; `persona` and `toolOrder` to `dsh-system-prompt`; `tools` to the tool registry for its presentation mode; `sessionTitle` to the fallback title service; `skills.registry`, `skills.local`, and `skills.tool` to the skill registry, local provider, and model-facing consumer; the required `workspaceContext` choice to `dsh-workspace-context` (`{ maxBytes }` enables loading and `false` disables it); `invariants` to the invariant service; and `toolBash`/`toolTasks` to the two model-facing tool plugins the bundle owns. It always mounts `dsh-llm-retry`, while each leaf adapter owns its nested `retryPolicy`. Omitted `sessionTitle` uses the explicit example policy of 5 words, 40 fallback bytes, and 80 accepted-title bytes. A `goals` object opts into the persisted domain, model tools, and same-session driver while forwarding `goals.domain` and `goals.tool` to their owners; omission or `false` leaves the stack absent so headless callers retain one-turn settlement. Set `skills.enabled: false` to omit both the local provider and model-facing skill tool, and set `toolTasks: false` to retain the task service for foreground producers without exposing `task_output` / `task_list` / `task_kill`. It resolves `dshHome` once through [`@deepseek-ai/dsh-paths`](../../util/paths/README.md) and forwards that absolute value to tool-bash's managed environment and enabled local skill discovery. An absent top-level `dshHome` adopts `skills.local.dshHome`; supplying both with different resolved paths fails loudly. `toolBash.enableRunInBackground` controls only the bash producer; independently loaded producers keep their own config. Workspace instructions register before the skill catalog so their session-prefix message renders first. App packages use `pickSpineConfig()` to copy only these bundle-owned fields.

 For example, `{ invariants: { enabled: true, package_allowlist: ['^@deepseek-ai/dsh-'], package_blocklist: ['agent-loop$'] } }` keeps the package-owned companions mounted but suppresses the blocked owner. Blocklist matches override allowlist matches; see [`dsh-invariants`](../../support/invariants/README.md) for regex and lifecycle rules.

@@ -67,7 +67,7 @@ For example, `{ invariants: { enabled: true, package_allowlist: ['^@deepseek-ai/

 A YAML include can deduplicate config but cannot own a bin or provide front-door defaults. The ACP app package makes protocol-pure stdout wiring the default, though a leaf can still add an unsafe logger. Bundle children register services in the root isolate-keyed store, so injected leaf siblings see them without load-order coupling.

-The bounded retry policy may repeat a transiently failed request in a new numbered step. Retry status and failed partial chunks stay outside model history, each provider attempt can still incur billing, front doors derive usage across every logged step, and the reconstructed request preserves the prior prefix for provider cache reuse.
+The retry policy may repeat a failed request in a new numbered step. Retry status, provider errors, and failed partial chunks stay outside model history; each provider attempt can still incur billing, always mode has no attempt limit, front doors derive usage across every logged step, and the reconstructed request preserves the prior prefix for provider cache reuse.

 ## Model Experience

--- a/packages/examples/agent-spine-demo/README.zh.md
+++ b/packages/examples/agent-spine-demo/README.zh.md
@@ -23,7 +23,7 @@
@deepseek-ai/dsh-goal             optional persisted same-session goal domain
@deepseek-ai/dsh-tool-goal        optional model-facing goal controls
@deepseek-ai/dsh-goal-session     optional same-session goal-round driver
-@deepseek-ai/dsh-llm-retry        bounded transient request retry policy
+@deepseek-ai/dsh-llm-retry        provider-routed request retry policy
@deepseek-ai/dsh-tasks-local      generic background-task registry
@deepseek-ai/dsh-invariants       configurable invariant registry service
@deepseek-ai/dsh-session/invariant
@@ -55,11 +55,11 @@

 ```ts
 import type { Config } from '@deepseek-ai/dsh-agent-spine-demo'
-// { agents?, maxParallelToolCalls?, persona?, toolOrder?, tools?, dshHome?, sessionTitle?, skills?, workspaceContext, toolBash?, toolTasks?, goals?, invariants?, llmRetry? }
+// { agents?, maxParallelToolCalls?, persona?, toolOrder?, tools?, dshHome?, sessionTitle?, skills?, workspaceContext, toolBash?, toolTasks?, goals?, invariants? }
 // workspaceContext requires { maxBytes } or false; the other owner schemas supply defaults.
 ```

-组合包将每个字段转发给拥有它的子节点：`agents` 与 `maxParallelToolCalls` 交给 `agent-loop`（`agents` 默认为 `[]`，上限在该处默认），因此每个应用提供自己的预创建 agent；TUI 和无头应用预创建 `main`，ACP 应用则在 `session/new` 按需创建 agent；`llmRetry` 交给有界重试策略；`persona` 与 `toolOrder` 交给 `dsh-system-prompt`；`tools` 交给工具注册表以配置呈现 mode；`sessionTitle` 交给后备标题服务；`skills.registry`、`skills.local` 与 `skills.tool` 分别交给 skill 注册表、本地提供方和面向模型的消费方；必填的 `workspaceContext` 选择交给 `dsh-workspace-context`（`{ maxBytes }` 启用加载，`false` 禁用）；`invariants` 交给不变式服务；`toolBash`/`toolTasks` 交给组合包拥有的两个面向模型工具插件。省略 `sessionTitle` 时采用显式示例策略：5 个词、40 个后备字节、80 个可接受标题字节。`goals` 对象会选用持久领域、模型工具和同会话驱动器，并将 `goals.domain` 与 `goals.tool` 转发给各自拥有者；省略或设为 `false` 会让整个栈缺席，使无头调用方继续以一轮结算。设置 `skills.enabled: false` 会同时省略本地提供方和面向模型的 skill 工具；设置 `toolTasks: false` 会保留供前台生产方使用的任务服务，但不公开 `task_output`/`task_list`/`task_kill`。它对 `dshHome` 只解析一次，解析通过 [`@deepseek-ai/dsh-paths`](../../util/paths/README.md) 完成，并将所得绝对值转发给 tool-bash 的托管环境和已启用的本地 skill 发现。顶层 `dshHome` 缺席时采用 `skills.local.dshHome`；两者同时提供但解析后的路径不同会明确失败。`toolBash.enableRunInBackground` 只控制 bash 生产方；独立加载的生产方保留各自配置。Workspace 指令先于 skill 目录注册，因此其会话前缀消息先渲染。应用包使用 `pickSpineConfig()`，只复制这些由组合包拥有的字段。
+组合包将每个字段转发给拥有它的子节点：`agents` 与 `maxParallelToolCalls` 交给 `agent-loop`（`agents` 默认为 `[]`，上限在该处默认），因此每个应用提供自己的预创建 agent；TUI 和无头应用预创建 `main`，ACP 应用则在 `session/new` 按需创建 agent；`persona` 与 `toolOrder` 交给 `dsh-system-prompt`；`tools` 交给工具注册表以配置呈现 mode；`sessionTitle` 交给后备标题服务；`skills.registry`、`skills.local` 与 `skills.tool` 分别交给 skill 注册表、本地提供方和面向模型的消费方；必填的 `workspaceContext` 选择交给 `dsh-workspace-context`（`{ maxBytes }` 启用加载，`false` 禁用）；`invariants` 交给不变式服务；`toolBash`/`toolTasks` 交给组合包拥有的两个面向模型工具插件。组合包始终挂载 `dsh-llm-retry`，而每个叶节点适配器拥有自己的嵌套 `retryPolicy`。省略 `sessionTitle` 时采用显式示例策略：5 个词、40 个后备字节、80 个可接受标题字节。`goals` 对象会选用持久领域、模型工具和同会话驱动器，并将 `goals.domain` 与 `goals.tool` 转发给各自拥有者；省略或设为 `false` 会让整个栈缺席，使无头调用方继续以一轮结算。设置 `skills.enabled: false` 会同时省略本地提供方和面向模型的 skill 工具；设置 `toolTasks: false` 会保留供前台生产方使用的任务服务，但不公开 `task_output`/`task_list`/`task_kill`。它对 `dshHome` 只解析一次，解析通过 [`@deepseek-ai/dsh-paths`](../../util/paths/README.md) 完成，并将所得绝对值转发给 tool-bash 的托管环境和已启用的本地 skill 发现。顶层 `dshHome` 缺席时采用 `skills.local.dshHome`；两者同时提供但解析后的路径不同会明确失败。`toolBash.enableRunInBackground` 只控制 bash 生产方；独立加载的生产方保留各自配置。Workspace 指令先于 skill 目录注册，因此其会话前缀消息先渲染。应用包使用 `pickSpineConfig()`，只复制这些由组合包拥有的字段。

 例如，`{ invariants: { enabled: true, package_allowlist: ['^@deepseek-ai/dsh-'], package_blocklist: ['agent-loop$'] } }` 会让包拥有的配套插件保持挂载，但抑制被阻止的拥有者。Blocklist 匹配优先于 allowlist 匹配；正则表达式与生命周期规则见 [`dsh-invariants`](../../support/invariants/README.md)。

@@ -67,7 +67,7 @@ import type { Config } from '@deepseek-ai/dsh-agent-spine-demo'

 YAML include 可以去重配置，却无法拥有 bin 或提供前端入口默认值。ACP 应用包默认接出协议纯净的 stdout，但叶节点仍可添加不安全的 logger。组合包子节点把服务注册到根 isolate-keyed store，因此注入这些服务的叶节点同级插件无需依赖加载顺序即可看到它们。

-有界重试策略可能在新的编号步骤中重复瞬时失败的请求。重试状态和失败的部分 chunk 不进入模型历史；每次提供方尝试仍可能产生计费；前端入口从所有已记录步骤推导用量；重建的请求保留先前前缀，以便复用提供方 cache。
+重试策略可能在新的编号步骤中重复失败的请求。重试状态、提供方错误和失败的部分 chunk 不进入模型历史；每次提供方尝试仍可能产生计费；always mode 没有尝试次数上限；前端入口从所有已记录步骤推导用量；重建的请求保留先前前缀，以便复用提供方 cache。

 ## 模型体验

--- a/packages/examples/agent-spine-demo/package.json
+++ b/packages/examples/agent-spine-demo/package.json
@@ -1,6 +1,6 @@
 {
  "name": "@deepseek-ai/dsh-agent-spine-demo",
-  "description": "The default executor-less/UI-less agent spine with fallback session titles, bounded retry, and optional persisted goals",
+  "description": "The default executor-less/UI-less agent spine with fallback session titles, provider-routed retry, and optional persisted goals",
  "version": "0.0.1",
  "private": true,
  "type": "module",
@@ -66,6 +66,7 @@
    "@deepseek-ai/dsh-llm": "workspace:^",
    "@deepseek-ai/dsh-paths": "workspace:^",
    "@deepseek-ai/dsh-llm-retry": "workspace:^",
+    "@deepseek-ai/dsh-subprocess-local": "workspace:^",
    "@deepseek-ai/dsh-sandbox-local": "workspace:^",
    "@deepseek-ai/dsh-sandbox-policy": "workspace:^",
    "@deepseek-ai/dsh-scope": "workspace:^",
--- a/packages/examples/agent-spine-demo/src/index.ts
+++ b/packages/examples/agent-spine-demo/src/index.ts
@@ -74,8 +74,9 @@ export interface GoalConfig {
 * `dshHome` to bash environment and local skill discovery, `sessionTitle` to
 * the fallback title service, `skills` to the
 * skill registry/local provider/tool consumer, `workspaceContext` to the
- * workspace-context loader, `llmRetry` to the bounded request-recovery policy,
- * and `toolBash`/`toolTasks` to the model-facing tool plugins this bundle owns.
+ * workspace-context loader, and `toolBash`/`toolTasks` to the model-facing tool
+ * plugins this bundle owns. Provider adapters own their `retryPolicy`; this
+ * bundle always mounts its executor.
 * `goals` opts into and configures the persisted goal domain plus its model tool
 * and same-session driver; `invariants` configures global and package-filtered
 * relational checks. Owner schemas supply defaults for optional input;
@@ -111,8 +112,6 @@ export interface Config {
  invariants?: InvariantConfig
  /** Opt-in persisted same-session goal stack; set false or omit to leave it unmounted. */
  goals?: GoalConfig | false
-  /** Bounded transient model-request retry policy. */
-  llmRetry?: llmRetry.Config
 }

 /** The skill config schema exported for app packages that forward `skills`. */
@@ -139,9 +138,6 @@ export const GoalConfigSchema: z<GoalConfig> = z.object({
  tool: toolGoal.Config,
 })

-/** The bounded LLM retry schema exported for app packages that forward `llmRetry`. */
-export const LlmRetryConfigSchema: z<llmRetry.Config> = llmRetry.Config
-
 /** Intersect the owners' schemas so validation + defaulting stay identical. */
 export const Config = z.intersect([
  AgentLoop.Config,
@@ -156,8 +152,7 @@ export const Config = z.intersect([
    toolTasks: z.union([z.const(false), ToolTasksConfigSchema]),
    invariants: InvariantService.Config,
    goals: z.union([z.const(false), GoalConfigSchema]),
-    llmRetry: LlmRetryConfigSchema,
-  }) as unknown as z<Pick<Config, 'tools' | 'dshHome' | 'sessionTitle' | 'skills' | 'workspaceContext' | 'toolBash' | 'toolTasks' | 'invariants' | 'goals' | 'llmRetry'>>,
+  }) as unknown as z<Pick<Config, 'tools' | 'dshHome' | 'sessionTitle' | 'skills' | 'workspaceContext' | 'toolBash' | 'toolTasks' | 'invariants' | 'goals'>>,
 ]) as unknown as z<Config>

 /**
@@ -179,7 +174,6 @@ export function pickSpineConfig(config: Omit<Config, 'agents'>): Omit<Config, 'a
    ...config.toolTasks !== undefined ? { toolTasks: config.toolTasks } : {},
    ...config.invariants !== undefined ? { invariants: config.invariants } : {},
    ...config.goals !== undefined ? { goals: config.goals } : {},
-    ...config.llmRetry !== undefined ? { llmRetry: config.llmRetry } : {},
  }
 }

@@ -217,7 +211,7 @@ export function apply(ctx: Context, config: Config): void {
    ctx.plugin(SkillLocal, Object.assign({}, config.skills?.local, { dshHome }))
  }
  ctx.plugin(AgentRegistry)
-  ctx.plugin(llmRetry, config.llmRetry ?? {})
+  ctx.plugin(llmRetry)
  if (config.goals !== undefined && config.goals !== false) {
    ctx.plugin(GoalService, config.goals.domain ?? {})
    ctx.plugin(toolGoal, config.goals.tool ?? {})
--- a/packages/examples/agent-spine-demo/tests/agent-core.spec.ts
+++ b/packages/examples/agent-spine-demo/tests/agent-core.spec.ts
@@ -12,7 +12,16 @@ import LocalBashExecutor from '@deepseek-ai/dsh-bash-local'
 import LocalFileSystem from '@deepseek-ai/dsh-fs-local'
 import * as ToolFs from '@deepseek-ai/dsh-tool-fs'
 import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
-import { CallId, LlmAdapter, LlmError, type GenerateOptions, type Message, type StreamChunk } from '@deepseek-ai/dsh-llm'
+import {
+  CallId,
+  LlmAdapter,
+  LlmError,
+  resolveRetryPolicy,
+  type GenerateOptions,
+  type Message,
+  type ResolvedRetryPolicy,
+  type StreamChunk,
+} from '@deepseek-ai/dsh-llm'
 import type { ToolExecution } from '@deepseek-ai/dsh-tools'
 import * as sessionInvariant from '@deepseek-ai/dsh-session/invariant'
 import * as agentInvariant from '@deepseek-ai/dsh-agent/invariant'
@@ -28,12 +37,9 @@ declare module '@deepseek-ai/dsh-tasks' {
 }

 async function composePrefix(ctx: Context, cwd: string): Promise<Message[]> {
-  const agent = { session: { header: { cwd } } } as unknown as Agent
-  const empty: Message[] = []
-  return await agentEvents(ctx, agent).waterfall(
-    'agent/session-prefix', empty, new AbortController().signal,
-    () => Promise.resolve(empty),
-  )
+  const agent = ctx.agentLoop.create(SessionId('agent-spine-prefix'), {}, { cwd })
+  await agentEvents(ctx, agent).serial('agent/step', 1, 1, new AbortController().signal)
+  return agent.session.deriveMessages()
 }

 /**
@@ -101,15 +107,8 @@ async function withIsolatedSkillHomes<T>(run: () => Promise<T>): Promise<T> {
  }
 }

-function waitForIdle(ctx: Context, target: Agent): Promise<void> {
-  return new Promise((resolve) => {
-    const dispose = ctx.on('agent/status', (agent, status) => {
-      if (agent === target && status === 'idle') {
-        dispose()
-        resolve()
-      }
-    })
-  })
+function waitForIdle(_ctx: Context, target: Agent): Promise<void> {
+  return target.whenIdle()
 }

 function messageText(message: Message | undefined): string {
@@ -118,6 +117,15 @@ function messageText(message: Message | undefined): string {

 class TransientOnceAdapter extends LlmAdapter {
  requests = 0
+  private readonly retryPolicy = resolveRetryPolicy({
+    mode: 'normal',
+    maxRetries: 1,
+    backoff: { initialDelayMs: 1, maxDelayMs: 1, jitterRatio: 0 },
+  }, 'agent-spine test provider retryPolicy')
+
+  override providerRetryPolicy(_provider: string): ResolvedRetryPolicy {
+    return this.retryPolicy
+  }

  async * stream(_options: GenerateOptions): AsyncIterable<StreamChunk> {
    this.requests += 1
@@ -221,15 +229,7 @@ describe('dsh-agent-spine-demo bundle', () => {

  it('loads and configures bounded request recovery for every bundled front door', async () => {
    const adapter = new TransientOnceAdapter()
-    const ctx = await mount({
-      workspaceContext: false,
-      llmRetry: {
-        maxTransientRetries: 1,
-        initialDelayMs: 1,
-        maxDelayMs: 1,
-        jitterRatio: 0,
-      },
-    })
+    const ctx = await mount({ workspaceContext: false })
    ctx.llm.registerAdapter(['mock'], adapter)
    const handle = await ctx.agents.create({
      sessionId: SessionId('bundled-retry-session'),
@@ -237,14 +237,14 @@ describe('dsh-agent-spine-demo bundle', () => {
      agentOptions: { provider: 'mock', model: 'mock' },
    })

-    handle.agent.followup([{ type: 'text', text: 'recover' }])
+    handle.agent.followup({ content: [{ type: 'text', text: 'recover' }], source: { kind: 'user' } })
    await waitForIdle(ctx, handle.agent)

    expect(adapter.requests).toBe(2)
    const retryEvents = handle.agent.session.events.filter(event => event.type === 'llm/retry')
    expect(retryEvents).toHaveLength(1)
    expect(retryEvents[0]?.data.retry).toBe(1)
-    expect(retryEvents[0]?.data.maxRetries).toBe(1)
+    expect(retryEvents[0]?.data).toMatchObject({ provider: 'mock', mode: 'normal', maxRetries: 1 })
    expect(handle.agent.session.events.find(event => event.type === 'session/title')?.data.title).toBe('recover')
    expect(messageText(handle.agent.session.deriveMessages().at(-1))).toBe('recovered by bundled policy')
    await handle.dispose()
@@ -337,7 +337,7 @@ describe('dsh-agent-spine-demo bundle', () => {
      })
      const agent = handle.agent

-      agent.followup([{ type: 'text', text: 'hi' }])
+      agent.followup({ content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } })
      await waitForIdle(ctx, agent)

      const sentText = adapter.requests[0]?.messages.map(messageText).join('\n')
@@ -366,7 +366,7 @@ describe('dsh-agent-spine-demo bundle', () => {
        agentOptions: { provider: 'mock', model: 'mock' },
      })

-      handle.agent.followup([{ type: 'text', text: 'hi' }])
+      handle.agent.followup({ content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } })
      await waitForIdle(ctx, handle.agent)

      expect(adapter.requests[0]?.messages).toEqual([{ role: 'user', content: [{ type: 'text', text: 'hi' }] }])
@@ -441,7 +441,10 @@ describe('dsh-agent-spine-demo bundle', () => {
        agentOptions: { provider: 'mock', model: 'mock' },
      })

-      handle.agent.followup([{ type: 'text', text: 'Create and load the project skill.' }])
+      handle.agent.followup({
+        content: [{ type: 'text', text: 'Create and load the project skill.' }],
+        source: { kind: 'user' },
+      })
      await waitForIdle(ctx, handle.agent)

      expect(adapter.requests).toHaveLength(4)
@@ -451,7 +454,7 @@ describe('dsh-agent-spine-demo bundle', () => {
          expect.not.stringContaining('hot-skill'),
        ])
      const catalogRequest = adapter.requests[2]?.messages.map(messageText).join('\n')
-      expect(catalogRequest).toContain('The available skill catalog changed.')
+      expect(catalogRequest).toContain('The following skills are available in this session:')
      expect(catalogRequest).toContain('- `hot-skill`: Hot-added skill')
      const loadedRequest = JSON.stringify(adapter.requests[3]?.messages)
      expect(loadedRequest).toContain('<skill_instructions>')
@@ -464,11 +467,6 @@ describe('dsh-agent-spine-demo bundle', () => {
          return [{
            type: event.type,
            source: event.data.source,
-            meta: {
-              kind: (event.data.meta as { kind?: unknown } | undefined)?.kind,
-              version: (event.data.meta as { version?: unknown } | undefined)?.version,
-              digest: typeof (event.data.meta as { digest?: unknown } | undefined)?.digest,
-            },
            text: event.data.content.map(block => block.type === 'text' ? block.text : '').join('\n'),
          }]
        }
@@ -496,23 +494,18 @@ describe('dsh-agent-spine-demo bundle', () => {
            "type": "tool/result",
          },
          {
-            "meta": {
-              "digest": "string",
-              "kind": "skill-catalog",
-              "version": 1,
-            },
            "source": {
              "kind": "plugin",
              "plugin": "tool-skill",
            },
            "text": "<system-reminder>
-        The available skill catalog changed. This complete catalog replaces every earlier available-skills list in this session:
+        A skill is a reusable set of task-specific instructions. The following skills are available in this session:

        <available_skills>
        - \`hot-skill\`: Hot-added skill
        </available_skills>

-        Use only names in this replacement catalog. If the user names a listed skill, or the task clearly matches its description, call the \`skill\` tool with the exact name before acting.
+        If the user names a skill, or the task clearly matches a skill's description, call the \`skill\` tool with the exact skill name before taking task actions. Load all applicable skills, then follow their full instructions. This catalog contains summaries only; do not infer or follow a skill's instructions until it has been loaded.
        </system-reminder>",
            "type": "user/message",
          },
@@ -597,11 +590,11 @@ describe('dsh-agent-spine-demo bundle', () => {
        agentOptions: { provider: 'mock', model: 'mock' },
      })

-      handle.agent.followup([{ type: 'text', text: 'hi' }])
+      handle.agent.followup({ content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } })
      await waitForIdle(ctx, handle.agent)

-      expect(messageText(adapter.requests[0]?.messages[0])).toContain('workspace rule before skills')
-      expect(messageText(adapter.requests[0]?.messages[1])).toContain('prefix-order-skill')
+      expect(messageText(adapter.requests[0]?.messages[1])).toContain('workspace rule before skills')
+      expect(messageText(adapter.requests[0]?.messages[2])).toContain('prefix-order-skill')
      await handle.dispose()
      await ctx.fiber.dispose()
    } finally {
@@ -666,7 +659,6 @@ describe('dsh-agent-spine-demo bundle', () => {
      toolBash: { enableRunInBackground: false },
      toolTasks: false as const,
      invariants: { enabled: false },
-      llmRetry: { maxTransientRetries: 1, jitterRatio: 0 },
    }

    expect(agentCore.pickSpineConfig(appConfig)).toEqual({
@@ -680,7 +672,6 @@ describe('dsh-agent-spine-demo bundle', () => {
      toolBash: appConfig.toolBash,
      toolTasks: appConfig.toolTasks,
      invariants: appConfig.invariants,
-      llmRetry: appConfig.llmRetry,
    })
    expect(agentCore.pickSpineConfig({ workspaceContext: false })).toEqual({ workspaceContext: false })
  })
--- a/packages/examples/agent-spine-demo/tests/multi-project-sandbox.e2e.ts
+++ b/packages/examples/agent-spine-demo/tests/multi-project-sandbox.e2e.ts
@@ -5,6 +5,7 @@ import { basename, join } from 'node:path'
 import { afterEach, beforeEach, describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
 import { SandboxBashExecutor } from '@deepseek-ai/dsh-bash-sandbox'
+import LocalSubprocessService from '@deepseek-ai/dsh-subprocess-local'
 import * as FsPolicy from '@deepseek-ai/dsh-fs-policy'
 import SandboxedFileSystem from '@deepseek-ai/dsh-fs-sandbox'
 import { CallId } from '@deepseek-ai/dsh-llm'
@@ -52,6 +53,7 @@ beforeEach(async () => {
  ctx = new Context()
  await ctx.plugin(LocalSandboxProvider, {})
  await ctx.plugin(SandboxPolicyService, { mode: 'workspace-write', workspaceRoot: fallbackRoot })
+  await ctx.plugin(LocalSubprocessService)
  await ctx.plugin(SandboxBashExecutor, { cwd: fallbackRoot, timeoutMs: 30_000 })
  await ctx.plugin(SandboxedFileSystem, { cwd: fallbackRoot })
  await ctx.plugin(agentSpine, {
--- a/packages/examples/cli-demo/README.i18n.yaml
+++ b/packages/examples/cli-demo/README.i18n.yaml
@@ -1,6 +1,6 @@
 # Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
-#   pnpm run verify-translation-pairing --write
-README.md: 4e8e5388e17ab2879582286adf593c72fb2cf78f
-README.zh.md: 3cad72071184403a78b7906d637bba67bea1a64a
+#   pnpm run verify-translation-pairing --write packages/examples/cli-demo/README.md
+README.md: b8f2bde962738a1a23f0e57218ab0f90e8e0b705
+README.zh.md: 322ba3fb3b253d867832534bd33f65df3a5b8d37
--- a/packages/examples/cli-demo/README.md
+++ b/packages/examples/cli-demo/README.md
@@ -21,7 +21,6 @@ The package mounts no console logger, interactive UI, user-interaction service,
 | `skills` | owner defaults | skill registry, local provider, and model-facing skill tool |
 | `toolBash` | owner defaults | model-facing bash config, including this producer's background opt-in |
 | `toolTasks` | owner defaults | generic `task_output` wait bounds |
-| `llmRetry` | owner defaults | bounded transient model-request retry policy |
 | `persistenceRoot` | `./.sessions` | JSONL session root |
 | `persistenceCompression` | `'zstd'` | JSONL artifact encoding (`'zstd'` or raw `'none'`) |
 | `workspaceContext` | required | workspace-instruction byte budget, or `false` to disable loading |
--- a/packages/examples/cli-demo/README.zh.md
+++ b/packages/examples/cli-demo/README.zh.md
@@ -21,7 +21,6 @@
 | `skills` | 拥有者默认值 | Skill 注册表、本地提供方和面向模型的 skill 工具 |
 | `toolBash` | 拥有者默认值 | 面向模型的 bash 配置，包括此生产方对后台任务的选用 |
 | `toolTasks` | 拥有者默认值 | 通用 `task_output` 等待边界 |
-| `llmRetry` | 拥有者默认值 | 有界的瞬时模型请求重试策略 |
 | `persistenceRoot` | `./.sessions` | JSONL 会话根目录 |
 | `persistenceCompression` | `'zstd'` | JSONL 工件编码（`'zstd'` 或原始 `'none'`） |
 | `workspaceContext` | 必填 | Workspace 指令字节预算，或以 `false` 禁用加载 |
--- a/packages/examples/cli-demo/src/cli.ts
+++ b/packages/examples/cli-demo/src/cli.ts
@@ -225,22 +225,28 @@ export async function runOneShot(ctx: Context, options: OneShotOptions): Promise
  let targetTurn: number | undefined
  let reason: TurnEndReason | undefined
  let result = ''
-  const usageByStep = new Map<number, TokenUsage>()
+  const usageByStep = new Map<string, TokenUsage>()
  let outputError: Error | undefined
  let resolveTurn!: () => void
  let rejectTurn!: (error: Error) => void
-  let settled = false
+  let firstTurnEnded = false
  const turnEnded = new Promise<void>((resolve, reject) => {
    resolveTurn = resolve
    rejectTurn = reject
  })

  const settleResolved = (): void => {
-    settled = true
+    if (firstTurnEnded) return
+    firstTurnEnded = true
    resolveTurn()
  }
  const settleRejected = (error: Error): void => {
-    settled = true
+    // The once-registered abort listener is the only rejecter, and a settled
+    // prompt makes targetTurn defined so onAbort skips rejection entirely;
+    // kept for symmetry with settleResolved.
+    /* v8 ignore next -- unreachable second settlement, see above */
+    if (firstTurnEnded) return
+    firstTurnEnded = true
    rejectTurn(error)
  }
  const observe = (sessionId: string, event: SessionEvent): void => {
@@ -254,20 +260,26 @@ export async function runOneShot(ctx: Context, options: OneShotOptions): Promise
  }

  const disposeListener = ctx.on('session/event', (session, event) => {
-    if (session !== agent.session || settled) return
+    if (session !== agent.session) return
    if (targetTurn === undefined) {
      if (event.type !== 'turn/start' || event.data.trigger.kind !== 'message') return
      targetTurn = event.data.turn
+    } else if (event.type === 'turn/start' && event.data.trigger.kind === 'retry'
+      && reason?.kind === 'error') {
+      targetTurn = event.data.turn
+      reason = undefined
    }
    observe(session.id, event)
    if (event.type === 'assistant/chunk'
      && event.data.turn === targetTurn
      && event.data.chunk.type === 'usage') {
-      usageByStep.set(event.data.step, event.data.chunk.usage)
+      usageByStep.set(`${event.data.turn}/${event.data.step}`, event.data.chunk.usage)
    }
    if (event.type === 'assistant/message' && event.data.turn === targetTurn) {
      result = assistantText(event) ?? result
-      if (event.data.usage !== undefined) usageByStep.set(event.data.step, event.data.usage)
+      if (event.data.usage !== undefined) {
+        usageByStep.set(`${event.data.turn}/${event.data.step}`, event.data.usage)
+      }
    }
    if (event.type === 'turn/end' && event.data.turn === targetTurn) {
      reason = event.data.reason
@@ -289,14 +301,14 @@ export async function runOneShot(ctx: Context, options: OneShotOptions): Promise

  try {
    /* v8 ignore next -- skips send only when cancellation wins the listener-registration race above */
-    if (!settled) { // eslint-disable-line @typescript-eslint/no-unnecessary-condition
-      agent.followup([{ type: 'text', text: options.task }])
+    if (!firstTurnEnded) { // eslint-disable-line @typescript-eslint/no-unnecessary-condition
+      agent.followup({ content: [{ type: 'text', text: options.task }], source: { kind: 'user' } })
    }
    await turnEnded
  } finally {
    if (onAbort !== undefined) signal?.removeEventListener('abort', onAbort)
-    disposeListener()
    await agent.whenIdle()
+    disposeListener()
  }

  /* v8 ignore next 3 -- turnEnded resolves only from the matching branch that assigns both values */
@@ -380,7 +392,6 @@ export function formatTurnFailure(reason: TurnEndReason): string {
    case 'error': return `failed at step ${reason.step}: ${'failure' in reason ? reason.failure.message : reason.message}`
    case 'disposed': return 'was disposed'
    case 'max-tokens': return 'reached the model output-token limit'
-    case 'rejected': return `was rejected: ${reason.reason}`
    case 'interrupted': return 'was interrupted during persistence recovery'
    default: return `ended with ${JSON.stringify(reason)}`
  }
--- a/packages/examples/cli-demo/src/index.ts
+++ b/packages/examples/cli-demo/src/index.ts
@@ -50,8 +50,6 @@ export interface Config {
  toolBash?: NonNullable<agentCore.Config['toolBash']>
  /** Generic background-task control-tool config forwarded through agent-spine-demo. */
  toolTasks?: NonNullable<agentCore.Config['toolTasks']>
-  /** Bounded transient model-request retry policy forwarded through agent-spine-demo. */
-  llmRetry?: NonNullable<agentCore.Config['llmRetry']>
  /** Controls automatic AGENTS.md/CLAUDE.md loading; configure a byte budget or set `false`. */
  workspaceContext: agentCore.Config['workspaceContext']
 }
@@ -74,7 +72,6 @@ export const Config: z<Config> = z.object({
  tools: ToolRegistry.Config,
  toolBash: agentCore.ToolBashConfigSchema,
  toolTasks: z.union([z.const(false), agentCore.ToolTasksConfigSchema]),
-  llmRetry: agentCore.LlmRetryConfigSchema,
  workspaceContext: z.union([z.const(false), workspaceContext.Config]).required(),
 })
 /* jscpd:ignore-end */
--- a/packages/examples/cli-demo/tests/built-bin.e2e.ts
+++ b/packages/examples/cli-demo/tests/built-bin.e2e.ts
@@ -1,4 +1,3 @@
-import { spawn } from 'node:child_process'
 import { existsSync } from 'node:fs'
 import { mkdtemp, mkdir, readFile, readdir, rm, symlink, writeFile } from 'node:fs/promises'
 import { tmpdir } from 'node:os'
@@ -6,6 +5,7 @@ import { dirname, join } from 'node:path'
 import { promisify } from 'node:util'
 import { fileURLToPath } from 'node:url'
 import { zstdDecompress } from 'node:zlib'
+import { execa } from 'execa'
 import { afterEach, describe, expect, it } from 'vitest'

 /**
@@ -23,7 +23,7 @@ const decompress = promisify(zstdDecompress)
 const dshPackages = [
  'examples/agent-spine-demo', 'examples/cli-demo', 'core/agent', 'core/session',
  'core/system-prompt', 'core/tools', 'core/agent-loop', 'llm/llm', 'bash/bash',
-  'bash/bash-local', 'bash/tool-bash', 'support/invariants', 'ui/app-boot',
+  'bash/bash-local', 'bash/tool-bash', 'subprocess/subprocess', 'subprocess/subprocess-local', 'support/invariants', 'ui/app-boot',
  'session-persistence/session-persistence', 'session-persistence/session-checkpoint-policy',
  'session-persistence/session-persistence-jsonl',
  'context/workspace-context',
@@ -80,6 +80,8 @@ async function makeConsumer(): Promise<string> {
  await writeFile(join(dir, 'cordis.yml'), [
    '- id: mock-llm',
    "  name: './mock-llm.ts'",
+    '- id: subprocess',
+    "  name: '@deepseek-ai/dsh-subprocess-local'",
    '- id: bash',
    "  name: '@deepseek-ai/dsh-bash-local'",
    '- id: cli-agent',
@@ -114,36 +116,34 @@ interface BinResult {
  readonly stderr: string
 }

-function runBuiltBin(cwd: string, args: readonly string[], interrupt?: NodeJS.Signals): Promise<BinResult> {
-  return new Promise((resolveResult, reject) => {
-    const child = spawn(process.execPath, [cliBin, ...args], {
-      cwd,
-      env: { ...process.env, DSH_HOME: join(cwd, '.dsh'), DSH_AGENTS_HOME: join(cwd, '.agents') },
-      stdio: ['ignore', 'pipe', 'pipe'],
-    })
-    let stdout = ''
-    let stderr = ''
+async function runBuiltBin(cwd: string, args: readonly string[], interrupt?: NodeJS.Signals): Promise<BinResult> {
+  const subprocess = execa(process.execPath, [cliBin, ...args], {
+    cwd,
+    env: { DSH_HOME: join(cwd, '.dsh'), DSH_AGENTS_HOME: join(cwd, '.agents') },
+    stdin: 'ignore',
+    timeout: 25_000,
+    killSignal: 'SIGKILL',
+    reject: false,
+    stripFinalNewline: false,
+  })
+  // Genuinely custom mid-stream logic: the signal cases deliver `interrupt`
+  // once the first streamed chunk proves the turn is in flight.
+  if (interrupt !== undefined) {
+    let streamed = ''
    let interrupted = false
-    child.stdout.setEncoding('utf8')
-    child.stdout.on('data', (chunk: string) => {
-      stdout += chunk
-      if (interrupt !== undefined && !interrupted && stdout.includes('assistant/chunk')) {
+    subprocess.stdout.on('data', (chunk: Buffer) => {
+      streamed += chunk.toString('utf8')
+      if (!interrupted && streamed.includes('assistant/chunk')) {
        interrupted = true
-        child.kill(interrupt)
+        subprocess.kill(interrupt)
      }
    })
-    child.stderr.setEncoding('utf8')
-    child.stderr.on('data', (chunk: string) => { stderr += chunk })
-    const timer = setTimeout(() => {
-      child.kill('SIGKILL')
-      reject(new Error(`built CLI did not exit. stdout:\n${stdout}\nstderr:\n${stderr}`))
-    }, 25_000)
-    child.once('error', (error) => { clearTimeout(timer); reject(error) })
-    child.once('exit', (code, signal) => {
-      clearTimeout(timer)
-      resolveResult({ code: code ?? -1, signal, stdout, stderr })
-    })
-  })
+  }
+  const result = await subprocess
+  if (result.timedOut) {
+    throw new Error(`built CLI did not exit. stdout:\n${result.stdout}\nstderr:\n${result.stderr}`)
+  }
+  return { code: result.exitCode ?? -1, signal: result.signal ?? null, stdout: result.stdout, stderr: result.stderr }
 }

 let consumer: string | undefined
--- a/packages/examples/cli-demo/tests/cli-demo.spec.ts
+++ b/packages/examples/cli-demo/tests/cli-demo.spec.ts
@@ -1,9 +1,11 @@
 import { mkdtemp } from 'node:fs/promises'
+import { randomUUID } from 'node:crypto'
 import { tmpdir } from 'node:os'
 import { join } from 'node:path'
 import { Context } from 'cordis'
 import Loader from '@cordisjs/plugin-loader'
-import { agentEvents, type Agent } from '@deepseek-ai/dsh-agent'
+import { agentEvents } from '@deepseek-ai/dsh-agent'
+import { SessionId } from '@deepseek-ai/dsh-session'
 import { CallId, type Message } from '@deepseek-ai/dsh-llm'
 import { TOOL_ORDER_REST } from '@deepseek-ai/dsh-system-prompt'
 import type { ToolExecution } from '@deepseek-ai/dsh-tools'
@@ -33,18 +35,16 @@ async function mount(config: cliDemo.Config, withBash = false): Promise<Context>
    })
  }
  contexts.push(ctx)
+  config.persistenceRoot ??= await mkdtemp(join(tmpdir(), 'dsh-cli-demo-persistence-'))
  await ctx.plugin(cliDemo, config)
  await new Promise(resolve => setTimeout(resolve, 80))
  return ctx
 }

 async function composePrefix(ctx: Context): Promise<Message[]> {
-  const agent = { session: { header: { cwd: '/tmp' } } } as unknown as Agent
-  const empty: Message[] = []
-  return await agentEvents(ctx, agent).waterfall(
-    'agent/session-prefix', empty, new AbortController().signal,
-    () => Promise.resolve(empty),
-  )
+  const agent = ctx.agentLoop.create(SessionId(`cli-demo-prefix-${randomUUID()}`), {}, { cwd: '/tmp' })
+  await agentEvents(ctx, agent).serial('agent/step', 1, 1, new AbortController().signal)
+  return agent.session.deriveMessages()
 }

 afterEach(async () => {
--- a/packages/examples/cli-demo/tests/cli.spec.ts
+++ b/packages/examples/cli-demo/tests/cli.spec.ts
@@ -3,7 +3,15 @@ import { tmpdir } from 'node:os'
 import { join, resolve } from 'node:path'
 import { Context } from 'cordis'
 import type { Agent } from '@deepseek-ai/dsh-agent'
-import { CallId, LlmAdapter, type GenerateOptions, type StreamChunk, type TokenUsage } from '@deepseek-ai/dsh-llm'
+import {
+  CallId,
+  LlmAdapter,
+  resolveRetryPolicy,
+  type GenerateOptions,
+  type ResolvedRetryPolicy,
+  type StreamChunk,
+  type TokenUsage,
+} from '@deepseek-ai/dsh-llm'
 import { SessionId, type SessionEvent, type TurnEndReason } from '@deepseek-ai/dsh-session'
 import { afterEach, describe, expect, it } from 'vitest'
 import * as cliDemo from '../src/index.ts'
@@ -20,11 +28,19 @@ type ScriptEntry = readonly StreamChunk[] | 'hang'
 class ScriptedAdapter extends LlmAdapter {
  readonly requests: GenerateOptions[] = []
  private cursor = 0
+  private readonly retryPolicy = resolveRetryPolicy({
+    mode: 'normal',
+    backoff: { initialDelayMs: 1, maxDelayMs: 1, jitterRatio: 0 },
+  }, 'cli test provider retryPolicy')

  constructor(private readonly script: readonly ScriptEntry[]) {
    super()
  }

+  override providerRetryPolicy(_provider: string): ResolvedRetryPolicy {
+    return this.retryPolicy
+  }
+
  async * stream(options: GenerateOptions): AsyncIterable<StreamChunk> {
    this.requests.push(options)
    const entry = this.script[this.cursor++]
@@ -107,7 +123,6 @@ async function harness(script: readonly ScriptEntry[]): Promise<Harness> {
    persistenceRoot: root,
    skills: { local: { dshHome: join(skillHome, '.dsh'), agentsHome: join(skillHome, '.agents') } },
    workspaceContext: false,
-    llmRetry: { initialDelayMs: 1, maxDelayMs: 1, jitterRatio: 0 },
  })
  await new Promise(resolve => setTimeout(resolve, 80))
  ctx.llm.registerAdapter(['mock'], new ScriptedAdapter(script))
@@ -319,7 +334,7 @@ describe('runOneShot and executeCli', () => {
    const { ctx, agent, persistenceRoot } = await harness([textResponse('final answer')])
    const output = await invoke(ctx, ['task'])
    expect(output).toEqual({ code: 0, stdout: 'final answer\n', stderr: '' })
-    expect(agent.status).toBe('disposed')
+    expect(agent.status).toBe('idle')
    const files = await readdir(persistenceRoot, { recursive: true })
    expect(files.some(file => file.endsWith('.jsonl.zstd'))).toBe(true)
  })
@@ -372,18 +387,20 @@ describe('runOneShot and executeCli', () => {
    ctx.on('agent/inbox/enqueue', (subject) => {
      if (subject !== agent || injected) return
      injected = true
-      agent.inject([{ type: 'text', text: 'startup injection' }], { source: { kind: 'plugin', plugin: 'test' } })
+      agent.inject({ content: [{ type: 'text', text: 'startup injection' }], source: { kind: 'plugin', plugin: 'test' } })
      other.append('turn/start', { turn: 1, trigger: { kind: 'injection', source: { kind: 'plugin', plugin: 'test' } } })
      other.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
    })
    const output = await invoke(ctx, ['--output-format', 'stream-json', 'task'])
    const lines = output.stdout.trimEnd().split('\n').map(line => JSON.parse(line) as Record<string, unknown>)
    const events = lines.slice(0, -1).map(line => line['event'] as SessionEvent)
-    expect(lines.at(-1)).toMatchObject({ type: 'result', success: true, turn: 2, result: 'streamed' })
-    expect(events[0]).toMatchObject({ type: 'turn/start', data: { turn: 2, trigger: { kind: 'message' } } })
-    expect(events.at(-1)).toMatchObject({ type: 'turn/end', data: { turn: 2 } })
+    expect(lines.at(-1)).toMatchObject({ type: 'result', success: true, turn: 1, result: 'streamed' })
+    expect(events[0]).toMatchObject({ type: 'turn/start', data: { turn: 1, trigger: { kind: 'message' } } })
+    expect(events.at(-1)).toMatchObject({ type: 'turn/end', data: { turn: 1 } })
    expect(lines.slice(0, -1).every(line => line['sessionId'] === agent.session.id)).toBe(true)
-    expect(events.some(event => event.type === 'user/message' && event.data.source.kind !== 'user')).toBe(false)
+    expect(events.some(event => event.type === 'user/message'
+      && event.data.source.kind === 'plugin'
+      && event.data.source.plugin === 'test')).toBe(false)
  })

  it('emits partial data and a diagnostic for non-completed turns', async () => {
@@ -409,7 +426,7 @@ describe('runOneShot and executeCli', () => {
    expect(JSON.parse(output.stdout)).toMatchObject({ success: false, reason: { kind: 'aborted' } })
    expect(output.code).toBe(1)
    expect(output.stderr).toContain('turn 1 was aborted')
-    expect(agent.status).toBe('disposed')
+    expect(agent.status).toBe('idle')
  })

  it('contains stream-writer failures, cancels, flushes, and returns the output error', async () => {
@@ -444,7 +461,7 @@ describe('runOneShot and executeCli', () => {
    expect(output.code).toBe(1)
    expect(output.stdout).toBe('')
    expect(output.stderr).toContain('stdout closed')
-    expect(final.agent.status).toBe('disposed')
+    expect(final.agent.status).toBe('idle')

    const disposal = await harness([textResponse('answer')])
    const disposalOutput = await invoke(disposal.ctx, ['task'], { failDispose: true })
@@ -471,7 +488,7 @@ describe('runOneShot and executeCli', () => {
    startup.ctx.on('session/event', (session, event) => {
      if (session === startup.agent.session && event.type === 'assistant/chunk') started()
    })
-    startup.agent.followup([{ type: 'text', text: 'first' }])
+    startup.agent.followup({ content: [{ type: 'text', text: 'first' }], source: { kind: 'user' } })
    await running
    const startupAbort = new AbortController()
    const waiting = runOneShot(startup.ctx, { task: 'second', signal: startupAbort.signal })
@@ -499,7 +516,6 @@ describe('formatTurnFailure', () => {
      [{ kind: 'error', step: 3, failure: { message: 'provider bad', code: 'SERVER' } }, 'failed at step 3: provider bad'],
      [{ kind: 'disposed' }, 'was disposed'],
      [{ kind: 'max-tokens' }, 'output-token limit'],
-      [{ kind: 'rejected', reason: 'policy' }, 'was rejected: policy'],
      [{ kind: 'interrupted' }, 'persistence recovery'],
    ]
    for (const [reason, expected] of cases) expect(formatTurnFailure(reason)).toContain(expected)
--- a/packages/examples/tui-demo/src/index.ts
+++ b/packages/examples/tui-demo/src/index.ts
@@ -131,6 +131,7 @@ export function composeTuiApp(ctx: Context, config: Config): void {
  ctx.plugin(SessionQuerySqlite, { path: join(persistenceRoot, 'session-query.db') })
  ctx.plugin(SessionReferenceService, config.sessionReferences ?? {})
  ctx.plugin(UserInteractionService)
+  ctx.plugin(uiTui.TuiPromptService)
  ctx.plugin(uiTui, {
    ...config.ui,
    ...config.welcome === undefined ? {} : { welcome: config.welcome },
--- a/packages/examples/tui-demo/tests/tui-agent.spec.ts
+++ b/packages/examples/tui-demo/tests/tui-agent.spec.ts
@@ -1,6 +1,6 @@
 import { describe, expect, it } from 'vitest'
 import { join } from 'node:path'
-import type { Context } from 'cordis'
+import { Context } from 'cordis'
 import Loader from '@cordisjs/plugin-loader'
 import { TOOL_ORDER_REST } from '@deepseek-ai/dsh-system-prompt'
 import * as tuiAgent from '../src/index.ts'
@@ -40,7 +40,7 @@ describe('dsh-tui-demo app', () => {
      },
      welcome: 'TUI ready',
      resumeCommand: 'dsh --resume {session}',
-      ui: { color: false, maxToolOutputLines: 3 },
+      ui: { theme: { color: false }, maxToolOutputLines: 3 },
      skills: { tool: { catalogDescriptionMaxLength: 8 } },
      toolBash: { enableRunInBackground: false },
      toolTasks: { waitTimeoutMs: 7, maxWaitTimeoutMs: 11 },
@@ -55,6 +55,7 @@ describe('dsh-tui-demo app', () => {
      'SessionQuerySqlite',
      'SessionReferenceService',
      'UserInteractionService',
+      'TuiPromptService',
      'ui-tui',
      'agent-spine-demo',
      'tool-ask-user',
@@ -67,15 +68,15 @@ describe('dsh-tui-demo app', () => {
      candidateLimit: 7,
      maxReferenceBytes: 1234,
    })
-    const tuiConfig = calls[7]?.config as { sessionId: string }
+    const tuiConfig = calls[8]?.config as { sessionId: string }
    expect(tuiConfig).toMatchObject({
      welcome: 'TUI ready',
      resumeCommand: 'dsh --resume {session}',
-      color: false,
+      theme: { color: false },
      maxToolOutputLines: 3,
    })
    expect(tuiConfig.sessionId).toMatch(/^main-session-[0-9a-f-]{36}$/)
-    const spineConfig = calls[8]?.config as {
+    const spineConfig = calls[9]?.config as {
      readonly agents: Array<Record<string, unknown>>
      readonly goals: Record<string, never>
      readonly maxParallelToolCalls: number
@@ -111,8 +112,8 @@ describe('dsh-tui-demo app', () => {
    expect(calls[2]?.config).toEqual({ root: './.sessions' })
    expect(calls[5]?.config).toEqual({})
    // No configured welcome forwards none: the TUI banner sweeps in without a subtitle.
-    expect(calls[7]?.config).toEqual({ sessionId: 'persisted-session' })
-    expect((calls[8]?.config as { agents: Array<Record<string, unknown>> }).agents[0]).toMatchObject({
+    expect(calls[8]?.config).toEqual({ sessionId: 'persisted-session' })
+    expect((calls[9]?.config as { agents: Array<Record<string, unknown>> }).agents[0]).toMatchObject({
      id: 'main',
      resumeSessionId: 'persisted-session',
    })
@@ -128,12 +129,12 @@ describe('dsh-tui-demo app', () => {
      workspaceContext: false,
    })

-    const tuiConfig = calls[6]?.config as { sessionId: string }
+    const tuiConfig = calls[7]?.config as { sessionId: string }
    expect(tuiConfig.sessionId).toMatch(/^main-session-[0-9a-f-]{36}$/)
-    expect((calls[7]?.config as { agents: Array<Record<string, unknown>> }).agents[0])
+    expect((calls[8]?.config as { agents: Array<Record<string, unknown>> }).agents[0])
      .toMatchObject({ sessionId: tuiConfig.sessionId })
    expect(calls.map(call => call.name)).not.toContain('command-goal')
-    expect(calls[7]?.config).toMatchObject({ goals: false })
+    expect(calls[8]?.config).toMatchObject({ goals: false })
  })

  it('has the namespace-plugin export shape so the Loader keeps its schema', () => {