fix(subagent): finalize Codex provider composition
This commit is contained in:
pku-xht
+2
-2
@@ -2,5 +2,5 @@
|
||||
# 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 .agents/notes/proposed/feature/2026-08-04-claude-code-and-codex-subagent-backends.md
|
||||
2026-08-04-claude-code-and-codex-subagent-backends.md: 37f45f844c9411af0467397272649533ed4d44cc
|
||||
2026-08-04-claude-code-and-codex-subagent-backends.zh.md: dcd7b90231bdaed47435c27deef413d20f0b7f28
|
||||
2026-08-04-claude-code-and-codex-subagent-backends.md: 3b9fd51632439da5b3c3fd9187de552d6c9ca5e2
|
||||
2026-08-04-claude-code-and-codex-subagent-backends.zh.md: 36be903640ad1c839d45ed1bf5e605f4e4d6e000
|
||||
|
||||
+10
-10
@@ -8,15 +8,15 @@ English | [中文](2026-08-04-claude-code-and-codex-subagent-backends.zh.md)
|
||||
|
||||
The named [`ctx.subagents`](../../implemented/feature/2026-06-21-subagent-capability-seam.md) registry lets a parent agent delegate work without knowing how the child runs, but the harness needs first-party routes to the real Codex and Claude Code products. A useful first version must hand either product one self-contained task, let it work in the parent Session's workspace, return a final answer or an explicit failure or cancellation, and leave no managed product process behind.
|
||||
|
||||
The product integrations must not become second owners for task text, cwd, cancellation, result settlement, or process trees. They must also prove the real assembled path in required keyless tests. A direct model HTTP request, product double, or hand-mounted plugin cannot show that the Loader, fixed tool, provider registration, official product protocol, native authentication shape, final answer, and teardown work together.
|
||||
The product integrations must not become second owners for task text, cwd, cancellation, result settlement, or process trees. Required keyless evidence therefore separates two facts: a real-product test proves the official protocol, native authentication shape, final answer, and teardown, while a Loader composition test proves that the public package and documented tool configuration load without starting the product. Direct model HTTP or a product double cannot replace the former; a hand-mounted plugin cannot replace the latter.
|
||||
|
||||
## Proposal
|
||||
|
||||
The harness provides two sibling one-shot providers behind two fixed model-facing tools. `subagent_codex` selects the `codex` provider, and `subagent_claude_code` selects the `claude-code` provider. Each tool accepts only a standalone text task and binds its provider at deployment time; product selection and background execution are not model arguments.
|
||||
The harness publishes two sibling one-shot providers as independently installable, opt-in packages. A user loads a provider and the existing common subagent tool in their own `cordis.yml`: `subagent_codex` binds `codex`, while `subagent_claude_code` binds `claude-code`. The shipped CLI dependency closure and base, Web, and headless configurations load neither provider. Each tool accepts only a standalone text task; product selection and background execution are not model arguments.
|
||||
|
||||
The Codex provider is implemented against Codex 0.146.0. The Claude Code provider remains unimplemented. This Note remains proposed until both siblings and their combined evidence are present.
|
||||
|
||||
Both providers report `inheritsParentContext: false`, advertise no optional start capabilities, and pass the parent Session cwd without copying the parent conversation. Every call creates a fresh product process and a non-resumable product conversation. The shared subagent service continues to own request resolution, lifecycle events, result settlement, and foreground collection; the shared subprocess service owns credential scrubbing, process-tree termination, and whole-tree exit observation.
|
||||
Both providers report `inheritsParentContext: false`, advertise no optional start capabilities, and pass the parent Session cwd without copying the parent conversation. Their documented tools disable background execution and use `maxDepth: 'provider-managed'`, leaving recursion policy with the out-of-process product instead of sending a limit the provider cannot enforce. Every call creates a fresh product process and a non-resumable product conversation. The shared subagent service continues to own request resolution, lifecycle events, result settlement, and foreground collection; the shared subprocess service owns credential scrubbing, process-tree termination, and whole-tree exit observation.
|
||||
|
||||
```text
|
||||
fixed tool → shared subagent service → product provider → official product process
|
||||
@@ -30,7 +30,7 @@ fixed tool → shared subagent service → product provider → official product
|
||||
| --- | --- | --- | --- |
|
||||
| Resolve | `dsh-tool-subagent` and `ctx.subagents` | Validate the product's text-only input and derive native startup parameters | Unsupported context or malformed input fails before a run is published |
|
||||
| Start | `dsh-subprocess` owns every acquired process tree | Reach the smallest native point at which the product conversation and process can both be controlled | `start()` publishes one existing `SubagentRun`, or cleans up and rejects |
|
||||
| Run | The product owns its native protocol facts; the holder owns their mapping | Submit exactly one task and derive one shared `completed`, `error`, or `aborted` result | The parent receives only a final answer or an explicit failure |
|
||||
| Run | The product owns its native protocol facts; the holder owns their mapping | Submit exactly one task and derive an existing shared stop reason; Codex uses `max-tokens` only for explicit context exhaustion | The parent receives only a final answer or an explicit failure |
|
||||
| Dispose | The foreground consumer requests release; `dsh-subprocess` proves exit | Close the native protocol and express any best-effort native cancellation | Disposal is idempotent and returns only after the whole process tree exits |
|
||||
|
||||
## Codex provider
|
||||
@@ -39,9 +39,9 @@ fixed tool → shared subagent service → product provider → official product
|
||||
|
||||
Before publication, the provider validates a non-empty text-only task, starts the managed app-server in the parent workspace, completes `initialize` → `initialized`, and creates an `ephemeral: true` thread. The published run owns exactly one `turn/start`; its thread and turn ids remain private and are never persisted in the parent Session.
|
||||
|
||||
`turn/completed` is the authoritative remote terminal fact. The latest nonblank `agentMessage` with `phase: "final_answer"` wins. When the product emits no explicit final phase, the latest message with `phase: null` is the compatibility fallback; commentary never replaces either answer. A completed turn without an answer, a failed or interrupted remote turn, malformed wire data, protocol closure, early process exit, or unknown server request becomes `error`. Local cancellation wins its race and remains `aborted`.
|
||||
`turn/completed` is the authoritative remote terminal fact. The latest nonblank `agentMessage` with `phase: "final_answer"` wins. When the product emits no explicit final phase, the latest message with `phase: null` is the compatibility fallback; commentary never replaces either answer. A failed turn with `error.codexErrorInfo: "contextWindowExceeded"` becomes `max-tokens`. A completed turn without an answer, every other failed or interrupted remote turn, malformed wire data, protocol closure, early process exit, or unknown server request becomes `error`; this version has no native refusal terminal and therefore produces no `refusal`. Local cancellation wins its race and remains `aborted`.
|
||||
|
||||
The unattended wire declines command and file approvals, grants no requested permissions for the turn, and declines MCP elicitation. Any other server request fails the run instead of waiting for a user interface the provider does not supply.
|
||||
For command and file approvals, the unattended wire selects a non-approval decision offered by the request, preferring `cancel`; the stable 0.146.0 request shape without an offered-decision list falls back to `decline`. It grants no requested permissions for the turn, answers user-input requests with no answers, and declines MCP elicitation. A request with no legal unattended response, or any unknown server request, fails the run instead of waiting for a user interface the provider does not supply.
|
||||
|
||||
An unpublished startup failure closes the wire, terminates the acquired process tree, waits for exit, and then rejects `start()`. Published disposal best-effort interrupts a known turn, closes the wire, ends stdin, invokes the shared termination escalation, and waits for whole-tree exit. Result failure and teardown failure stay independently observable.
|
||||
|
||||
@@ -51,11 +51,11 @@ The Claude Code sibling is not yet implemented. Its product version, official in
|
||||
|
||||
## Evidence contract
|
||||
|
||||
Each product owns branch-complete package tests, a required real-product spec, and a real Loader snapshot. The real-product tier uses the exact official distribution under test, a non-empty fake product key, an isolated temporary workspace and product home, and a loopback fixed-answer model. Missing product requests, wrong authentication, altered task text, a non-exact answer, a skipped real product, or a surviving managed handle fails the required test.
|
||||
Each product owns branch-complete package tests, a required real-product spec, and a Loader composition e2e. The real-product tier uses the exact official distribution under test, a non-empty fake product key, an isolated temporary workspace and product home, and a loopback fixed-answer model. Missing product requests, wrong authentication, altered task text, a non-exact answer, a skipped real product, or a surviving managed handle fails the required test. The separate Loader tier boots the README-shaped user configuration, verifies the fixed provider and foreground-only common tool, and must not start a product process.
|
||||
|
||||
The Codex evidence pins `@openai/codex@0.146.0` and `codex-cli 0.146.0`. Its real-product spec observes the exact Bearer key, original task, byte-exact final answer, unattended command rejection with no file side effect, local cancellation, and whole-tree exit. Its Loader snapshot fixes the no-background tool schema, exact tool call and result, complete persisted parent Session, product request, and pre-teardown quiescence. The npm package is a development dependency for reproducible evidence; production still supplies `codex` on `PATH`.
|
||||
The Codex evidence pins `@openai/codex@0.146.0` and `codex-cli 0.146.0`. Its real-product spec observes the exact Bearer key, original task, byte-exact final answer, unattended command rejection with no file side effect, local cancellation, and whole-tree exit. Its Loader e2e resolves `@deepseek-ai/dsh-subagent-codex` by package name, verifies the `codex` registration and `subagent_codex` schema with background omitted, accepts `maxDepth: 'provider-managed'`, and records zero child starts while no `codex` command is available. The npm package is a development dependency for reproducible real-product evidence; production still supplies `codex` on `PATH`.
|
||||
|
||||
The combined contract is complete only when the Claude sibling has equivalent real-product evidence and one assembled Loader run proves both fixed tools coexist without changing the common subagent contract.
|
||||
The combined contract is complete only when the Claude sibling has equivalent real-product evidence and both public Loader configurations prove the fixed tools use the unchanged common subagent contract.
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
@@ -73,7 +73,7 @@ The combined contract is complete only when the Claude sibling has equivalent re
|
||||
|
||||
## Acceptance criteria
|
||||
|
||||
Both fixed tools reach their corresponding real products through the Loader, return exact final answers or explicit failure or cancellation, persist the complete model-visible parent transcript, and prove managed process-tree quiescence in required keyless CI. Both packages document their configuration, lifecycle, failure behavior, model experience, and limitations; generated package, configuration, capability, dependency, and third-party records agree with the shipped manifests.
|
||||
Both public provider packages load from user-owned Cordis configurations and form their fixed foreground tools without appearing in the shipped CLI defaults. Separate required real-product specs return exact final answers or explicit failure or cancellation and prove managed process-tree quiescence. Both packages document their configuration, lifecycle, failure behavior, model experience, and limitations; generated package, configuration, capability, dependency, and third-party records agree with the shipped manifests.
|
||||
|
||||
The implemented Codex half satisfies this contract for its fixed tool and 0.146.0 baseline. The proposal becomes implemented only after the Claude Code sibling and the combined two-product evidence satisfy the same ownership and lifecycle boundaries.
|
||||
|
||||
|
||||
+10
-10
@@ -8,15 +8,15 @@ Status: proposed
|
||||
|
||||
命名的 [`ctx.subagents`](../../implemented/feature/2026-06-21-subagent-capability-seam.md) 注册表让父 agent（智能体）无需了解子级的运行方式即可委派工作，但 harness 需要通往真实 Codex 与 Claude Code 产品的第一方路径。可用的首版必须能向任一产品交付一项自包含任务，让它在父会话的工作区中执行，返回最终回答或明确的失败或取消结果，并且不留下任何受管的产品进程。
|
||||
|
||||
产品集成不得成为任务文本、cwd、取消、结果结算或进程树的第二责任方。它们还必须在强制性的无密钥测试中证明真实组装路径。直接发起模型 HTTP 请求、使用产品替身或手工挂载插件，都无法证明 Loader、固定工具、提供方注册、官方产品协议、原生身份验证形态、最终回答和资源清理能够协同工作。
|
||||
产品集成不得成为任务文本、cwd、取消、结果结算或进程树的第二责任方。因此，强制性的无密钥证据会分别证明两个事实：真实产品测试证明官方协议、原生身份验证形态、最终回答和资源清理；Loader 装配测试证明公开包与文档中的工具配置可以加载，且不会启动产品。直接发起模型 HTTP 请求或使用产品替身无法取代前者，手工挂载插件则无法取代后者。
|
||||
|
||||
## 提案
|
||||
|
||||
harness 在两个固定的面向模型工具背后提供两个一次性兄弟提供方。`subagent_codex` 选择 `codex` 提供方，`subagent_claude_code` 选择 `claude-code` 提供方。每个工具只接受独立文本任务，并在部署时绑定其提供方；产品选择与后台执行都不作为模型参数。
|
||||
harness 将两个一次性兄弟提供方发布为可独立安装的可选包。用户在自己的 `cordis.yml` 中加载提供方与现有的通用 subagent 工具：`subagent_codex` 绑定 `codex`，而 `subagent_claude_code` 绑定 `claude-code`。正式 CLI 的依赖闭包以及基础、Web 和 headless 配置都不加载这两个提供方。每个工具只接受独立文本任务；产品选择与后台执行都不作为模型参数。
|
||||
|
||||
Codex 提供方基于 Codex 0.146.0 实现。Claude Code 提供方仍未实现。在两个兄弟提供方及其组合证据全部具备之前，本 Agent Note 将保持提案状态。
|
||||
|
||||
这两个提供方都报告 `inheritsParentContext: false`，不声明任何可选的启动时功能，并传递父会话 cwd，但不会复制父级对话。每次调用都会创建一个全新的产品进程和一次不可续接的产品对话。共享 subagent 服务继续负责请求解析、生命周期事件、结果结算和前台收集；共享子进程服务负责凭证清洗、进程树终止以及整棵进程树的退出观测。
|
||||
这两个提供方都报告 `inheritsParentContext: false`，不声明任何可选的启动时功能，并传递父会话 cwd，但不会复制父级对话。文档中的工具会关闭后台执行并使用 `maxDepth: 'provider-managed'`，让进程外产品自行负责递归策略，而不会向提供方发送其无法执行的限制。每次调用都会创建一个全新的产品进程和一次不可续接的产品对话。共享 subagent 服务继续负责请求解析、生命周期事件、结果结算和前台收集；共享子进程服务负责凭证清洗、进程树终止以及整棵进程树的退出观测。
|
||||
|
||||
```text
|
||||
fixed tool → shared subagent service → product provider → official product process
|
||||
@@ -30,7 +30,7 @@ fixed tool → shared subagent service → product provider → official product
|
||||
| --- | --- | --- | --- |
|
||||
| 解析 | `dsh-tool-subagent` 与 `ctx.subagents` | 验证产品的纯文本输入并推导原生启动参数 | 不受支持的上下文或格式错误的输入会在发布运行前报错 |
|
||||
| 启动 | `dsh-subprocess` 负责每棵已获取的进程树 | 到达能够同时控制产品对话与进程的最小原生控制点 | `start()` 发布一个已存在的 `SubagentRun`，否则清理后拒绝调用 |
|
||||
| 运行 | 产品负责其原生协议事实；持有方负责映射这些事实 | 只提交一项任务，并推导一个共享的 `completed`、`error` 或 `aborted` 结果 | 父级只会收到最终回答或明确失败 |
|
||||
| 运行 | 产品负责其原生协议事实；持有方负责映射这些事实 | 只提交一项任务，并推导出一种现有的共享停止原因；Codex 仅在明确发生上下文耗尽时使用 `max-tokens` | 父级只会收到最终回答或明确失败 |
|
||||
| dispose（资源释放） | 前台消费方请求释放；`dsh-subprocess` 证明进程已退出 | 关闭原生协议，并发出尽力而为的原生取消请求 | 释放操作具有幂等性，且仅在整棵进程树退出后才返回 |
|
||||
|
||||
## Codex 提供方
|
||||
@@ -39,9 +39,9 @@ fixed tool → shared subagent service → product provider → official product
|
||||
|
||||
发布前，提供方会验证非空的纯文本任务，在父级工作区中启动受管的 app-server，完成 `initialize` → `initialized` 握手，并创建一个 `ephemeral: true` 线程。已发布的运行只拥有一次 `turn/start`；其线程 ID 与轮次 ID 保持私有，绝不会持久化到父会话。
|
||||
|
||||
`turn/completed` 是权威的远端终止事实。以最后一条非空白的 `agentMessage` 为准，但它必须带有 `phase: "final_answer"`。若产品没有发出明确的最终阶段，则以最后一条 `phase: null` 的消息作为兼容性回退；过程说明绝不会取代上述任一答案。轮次完成却没有答案、远端轮次失败或中断、协议数据格式错误、协议关闭、进程提前退出或未知的服务器请求，都会产生 `error`。本地取消在竞态中胜出并保持为 `aborted`。
|
||||
`turn/completed` 是权威的远端终止事实。以最后一条非空白的 `agentMessage` 为准，但它必须带有 `phase: "final_answer"`。若产品没有发出明确的最终阶段，则以最后一条 `phase: null` 的消息作为兼容性回退；过程说明绝不会取代上述任一答案。带有 `error.codexErrorInfo: "contextWindowExceeded"` 的失败轮次会成为 `max-tokens`。轮次完成却没有答案、其他任何远端失败或中断轮次、协议数据格式错误、协议关闭、进程提前退出或未知的服务器请求，都会产生 `error`；本版本没有原生的拒绝终止状态，因此不会产生 `refusal`。本地取消在竞态中胜出并保持为 `aborted`。
|
||||
|
||||
无人值守的协议连接会拒绝命令与文件审批，不授予该轮次请求的任何权限，并拒绝 MCP elicitation。其他任何服务器请求都会导致此次运行失败，而不会等待本提供方没有提供的用户界面。
|
||||
对于命令与文件审批，无人值守的协议连接会从请求给出的决策选项中选择一项不予批准的决策，并优先选择 `cancel`；稳定的 0.146.0 请求形态没有决策选项列表，因此回退到 `decline`。它不授予该轮次请求的任何权限，不向用户输入请求提供任何答案，并拒绝 MCP elicitation。若请求在无人值守模式下没有合法响应，或是未知服务器请求，此次运行就会失败，而不会等待本提供方没有提供的用户界面。
|
||||
|
||||
若启动在发布前失败，提供方会关闭协议连接、终止已获取的进程树并等待其退出，然后拒绝 `start()`。对已发布的运行执行释放时，提供方会尽力中断已知轮次、关闭协议连接、结束标准输入、调用共享的进程树逐级终止机制，并等待整棵进程树退出。结果失败与清理失败仍可彼此独立地观察。
|
||||
|
||||
@@ -51,11 +51,11 @@ Claude Code 兄弟提供方尚未实现。其中间提案不固定产品版本
|
||||
|
||||
## 证据契约
|
||||
|
||||
每个产品都负责覆盖所有分支的包（package）测试、一项必跑的真实产品测试和一个真实 Loader 快照。真实产品测试层级使用被测的确切官方发行版、非空的伪产品密钥、隔离的临时工作区与产品主目录，以及能返回固定答案的回环模型。产品请求缺失、身份验证错误、任务文本被改动、答案不完全一致、真实产品被跳过或受管句柄仍存活，都会使这项必跑测试失败。
|
||||
每个产品都负责覆盖所有分支的包（package）测试、一项必跑的真实产品测试和一项 Loader 装配 e2e。真实产品测试层级使用被测的确切官方发行版、非空的伪产品密钥、隔离的临时工作区与产品主目录，以及能返回固定答案的回环模型。产品请求缺失、身份验证错误、任务文本被改动、答案不完全一致、真实产品被跳过或受管句柄仍存活，都会使这项必跑测试失败。独立的 Loader 层级会启动与 README 同形的用户配置，验证固定提供方与只支持前台执行的通用工具，并且不得启动产品进程。
|
||||
|
||||
Codex 证据锁定 `@openai/codex@0.146.0` 与 `codex-cli 0.146.0`。其真实产品测试会观测确切的 Bearer 密钥、原始任务、逐字节完全一致的最终回答、不会产生文件副作用的无人值守命令拒绝、本地取消以及整棵进程树退出。其 Loader 快照锁定不支持后台执行的工具 schema、确切的工具调用与结果、完整的已持久化父会话、产品请求，以及清理前的完全停稳状态。该 NPM 包是用于复现证据的开发依赖；生产环境仍提供 `codex`，并通过 `PATH` 解析。
|
||||
Codex 证据锁定 `@openai/codex@0.146.0` 与 `codex-cli 0.146.0`。其真实产品测试会观测确切的 Bearer 密钥、原始任务、逐字节完全一致的最终回答、不会产生文件副作用的无人值守命令拒绝、本地取消以及整棵进程树退出。其 Loader e2e 会按包名解析 `@deepseek-ai/dsh-subagent-codex`，验证 `codex` 注册与省略后台参数的 `subagent_codex` schema，接受 `maxDepth: 'provider-managed'`，并在环境中没有可用 `codex` 命令时记录零次子级启动。该 NPM 包是用于复现真实产品证据的开发依赖；生产环境仍提供 `codex`，并通过 `PATH` 解析。
|
||||
|
||||
只有在 Claude 兄弟提供方具备同等的真实产品证据，并且一次组装后的 Loader 运行证明两个固定工具可以共存且无需更改通用 subagent 契约时，组合契约才算完整。
|
||||
只有在 Claude 兄弟提供方具备同等的真实产品证据，并且两个公开 Loader 配置都证明固定工具使用未变的通用 subagent 契约时，组合契约才算完整。
|
||||
|
||||
## 曾考虑的替代方案
|
||||
|
||||
@@ -73,7 +73,7 @@ Codex 证据锁定 `@openai/codex@0.146.0` 与 `codex-cli 0.146.0`。其真实
|
||||
|
||||
## 验收标准
|
||||
|
||||
两个固定工具都通过 Loader 到达相应的真实产品，返回完全一致的最终回答或明确的失败或取消结果，持久化完整的模型可见父级 transcript，并在强制性的无密钥 CI 中证明受管进程树完全停稳。两个包都会记录其配置、生命周期、失败行为、模型体验和限制；生成的包、配置、功能、依赖与第三方记录均与已交付的 manifest（元数据清单）一致。
|
||||
两个公开提供方包都能从用户自有的 Cordis 配置加载并组成固定的前台工具，而且不会出现在正式 CLI 默认配置中。独立的强制真实产品测试会返回完全一致的最终回答或明确的失败或取消结果，并证明受管进程树完全停稳。两个包都会记录其配置、生命周期、失败行为、模型体验和限制；生成的包、配置、功能、依赖与第三方记录均与已交付的 manifest（元数据清单）一致。
|
||||
|
||||
已经实现的 Codex 部分为其固定工具和 0.146.0 基线满足了本契约。只有在 Claude Code 兄弟提供方及两种产品的组合证据满足相同的归属与生命周期边界后，本提案才会进入 implemented 状态。
|
||||
|
||||
|
||||
Reference in New Issue
Block a user