maka-agent 0.2.1-dev.90.20261017 → 0.2.2-dev.101.20261026

package/DISCLAIMER-WIP

@@ -12,7 +12,7 @@ conditions. What follows is a list of issues the project is currently aware of
 (this list is likely to be incomplete):
 
  * The initial-code authorization review remains open at
-   https://github.com/apache/maka/issues/3268. Since RC1, the ASF has recorded
+   https://github.com/apache/maka/issues/3268. Since 0.2.0 RC1, the ASF has recorded
    the ICLA of contributor Mingqwqqaq, and the Deep Research implementation
    has been removed. The review still needs to record the authorization basis
    for the retained initial code; these changes do not establish that an SGA

package/LICENSE

@@ -218,7 +218,9 @@ License: Apache-2.0
 These patches modify the execution time policy and its SDK forwarding. The
 Apache License, Version 2.0 above applies. The generated run Worker includes
 embedded third-party code; its upstream notices and license texts are retained
-in patches/run-2.1.4-notices.md.
+in patches/run-2.1.4-notices.md. The patch carries no WebAssembly binary:
+postinstall substitutes the unchanged payload from the installed run package
+for `__MAKA_RUN_WASM_BASE64__` before applying the patch.
 
 Vercel AI SDK
 
@@ -240,7 +242,7 @@ License, Version 2.0 above applies to the adapted material.
 The pinned versions are the ones present when the adaptation was made. The
 adapted declarations are unchanged through @ai-sdk/provider-utils 5.0.25 and
 @ai-sdk/provider 4.0.7, the versions recorded by the code origin audit. Maka
-also redistributes a patch against @ai-sdk/provider-utils 5.0.40; that patch
+also redistributes a patch against @ai-sdk/provider-utils 5.0.51; that patch
 modifies Apache-2.0 source and is covered by the license reproduced above.
 
 Astryx
@@ -351,10 +353,10 @@ SOFTWARE.
 
 Zod dependency patch
 
-Source: https://www.npmjs.com/package/zod/v/4.6.5
+Source: https://www.npmjs.com/package/zod/v/4.6.7
 Repository: https://github.com/colinhacks/zod
-Version: 4.6.5
-Dependency patch: patches/zod+4.6.5.patch
+Version: 4.6.7
+Dependency patch: patches/zod+4.6.7.patch
 Copyright (c) 2025 Colin McDonnell
 License: MIT
 
@@ -385,12 +387,12 @@ SOFTWARE.
 
 Model Context Protocol client dependency patch
 
-Source: https://www.npmjs.com/package/@modelcontextprotocol/client/v/2.2.0
+Source: https://www.npmjs.com/package/@modelcontextprotocol/client/v/2.3.0
 Repository: https://github.com/modelcontextprotocol/typescript-sdk
-Version: 2.2.0
-Dependency patch: patches/@modelcontextprotocol+client+2.2.0.patch
+Version: 2.3.0
+Dependency patch: patches/@modelcontextprotocol+client+2.3.0.patch
 Copyright (c) 2024-2025 Model Context Protocol a Series of LF Projects, LLC.
-Declared package license: MIT
+Declared package license: Apache-2.0
 
 Maka redistributes a source patch that releases settled request observers under
 transport backpressure. The upstream package's LICENSE describes a transition
@@ -622,6 +624,9 @@ License: MIT
 adaptations of this component.
 The following MIT License applies to that material:
 
+Maka-authored modifications are licensed under the Apache License, Version 2.0
+reproduced above; the upstream portions retain the MIT terms below.
+
 MIT License
 
 Copyright (c) 2025 Cua AI, Inc.
@@ -809,8 +814,7 @@ Desktop renderer vendored source assets
 
 The source archive includes third-party SVG paths and source assets from
 Lobe Icons (MIT), LocalAI (MIT), Simple Icons (CC0-1.0), TDesign Icons (MIT),
-MingCute (Apache-2.0), Allogo (MIT), Ant Design Icons (MIT), and Semi Design
-(MIT), used by the desktop renderer. Their file-by-file inventory, immutable upstream revisions,
+MingCute (Apache-2.0), and Allogo (MIT), used by the desktop renderer. Their file-by-file inventory, immutable upstream revisions,
 copyright notices, and license texts are in:
 
   apps/desktop/src/renderer/public/THIRD_PARTY_LICENSES.txt

package/README.md

@@ -48,7 +48,7 @@ conditions. What follows is a list of issues the project is currently aware of
 (this list is likely to be incomplete):
 
  * The initial-code authorization review remains open at
-   https://github.com/apache/maka/issues/3268. Since RC1, the ASF has recorded
+   https://github.com/apache/maka/issues/3268. Since 0.2.0 RC1, the ASF has recorded
    the ICLA of contributor Mingqwqqaq, and the Deep Research implementation
    has been removed. The review still needs to record the authorization basis
    for the retained initial code; these changes do not establish that an SGA
@@ -142,6 +142,178 @@ Maka asks before privileged tool operations by default. `maka run --yolo` grants
 and network access and should only be used in an environment you are prepared to let the task
 modify.
 
+
+For a parent agent coordinating work across calls, prefer `--no-wait`, save the returned
+`sessionId` as `$session`, then wait when ready to handle an interaction or result:
+
+```sh
+maka run "Inspect this project" --no-wait
+maka run --session "$session" --wait --json
+```
+
+`--no-wait` already returns JSON; an extra `--json` is unnecessary. Ordinary `maka run` keeps its
+blocking default and prints the final answer, making it suitable for a single shell call.
+Use `maka run "Inspect this project" --json` for a structured blocking result.
+
+Choose the next Session call by purpose:
+
+```sh
+maka run --session "$session" --status
+maka run --session "$session" --recap --json
+maka run --session "$session" --answer '["Yes",null]' --json
+maka run --session "$session" "Inspect the tests next" --no-wait
+maka run --session "$session" "Focus on the parser" --steer --no-wait
+maka run --session "$session" --result --json
+maka run --session "$session" --resume --json
+```
+
+An added prompt queues by default, so `--queue` is optional. Repeating the same flag or value is
+accepted; conflicting values and incompatible actions still produce a usage error. Execution commands wait unless given
+`--no-wait`; `--session <id> --wait` explicitly observes the current Turn without submitting work.
+
+| Call | Default stdout | Meaning of exit 0 |
+| --- | --- | --- |
+| Ordinary run or Session wait/result | Final answer text; pending interactions are JSON | Inspect the outcome; a pending interaction still needs a decision |
+| `--no-wait` | JSON admission receipt | Submission accepted, not task completion |
+| `--status` / `--stop` | JSON | Observation or stop request succeeded |
+| Model settings without an action | JSON with `configuration` | Settings saved; execution is unchanged |
+| `--recap` | Summary text | Independent summary generated |
+
+Rejecting an approval informs the model and lets the same Turn continue within existing permissions
+or explain the blocker. A tool failure alone does not override the Runtime terminal status. Waits
+return at Turn completion or the next interaction; `--deny --no-wait` only submits the decision.
+
+Use `--json` on waits, results, and recaps for uniform structured results. `status` is the lifecycle;
+`finalOutput` is the task answer; `recap` is progress; `submission` is an admission receipt. Errors are brief text on stderr by default. Explicit `--json` returns structured errors on stdout,
+including the original message and Host `code` when available. An operation error does not assert
+that the Session failed. For a remote Host, repeat the same `--host <profile-id>`
+on every call. `--session` inherits saved settings and accepts persistent model/connection/thinking updates.
+Workspace, timeout, and Graph overrides still require legacy `--resume <id>` and are rejected
+before connecting when combined with `--session`.
+
+Change a Session's model persistently, then explicitly choose continuation or a new message:
+
+```sh
+maka run --session "$session" --model "other-model"
+maka run --session "$session" --model "other-model" --resume --json
+maka run --session "$session" --model "other-model" "Continue the review" --no-wait
+```
+
+These are three alternatives. The first only saves settings and returns a JSON `configuration`
+receipt with `connection`, `model`, and `thinking`; it does not start or resume execution.
+Add `--connection <slug>` to select a different saved account. Native native model/connection changes reset
+thinking to the model default unless `--thinking` is supplied; selecting the same target preserves the saved level. The new settings persist for future
+calls, like Desktop model selection; global defaults and permissions are unchanged.
+
+Settings are saved before an explicit action, so they remain saved if that later action fails.
+CLI does not classify why execution stopped or choose a model automatically. Host errors such
+as busy-session rejection are returned without implicitly stopping the Turn or resubmitting the
+old prompt. Resume uses the Host's existing safe-continuation operation; a new prompt is new input.
+
+`--recap` generates a one-sentence summary of recorded Session progress using the same
+independent model call as TUI `/recap`. It works during an active Turn without steering,
+queueing, stopping, or waiting for that Turn to finish. Without `--json` it prints the summary;
+with `--json` it returns `sessionId`, current `status`, and a separate `recap` field, for example:
+
+```json
+{"sessionId":"…","status":"running","queued":0,"recap":"You asked for a review; the parser is checked and test verification is still pending."}
+```
+
+The recap is a snapshot of the history captured when generation starts. It cannot see unrecorded
+work or results from a tool that has not returned, and execution can advance while it is generated.
+Each invocation requests a fresh summary and consumes model usage; `--status` remains the cheap
+status query. A recap is not the final answer (`--result`) or a reply from the running agent.
+Combine `--status --recap` to return JSON containing both status/boundary and progress. Other
+Session actions, prompts, and wait flags cannot be combined with recap.
+Successful recap generation exits 0 even if the Session's latest Turn failed; generation errors
+exit 1. Ctrl+C cancels the CLI observer and leaves Host execution running.
+
+Session controls (`--session`) inherit the Host's saved permissions, just as they inherit the
+model and working directory. Continuing, queueing, steering, or answering does not require another
+`--yolo` and does not change the boundary. Add `--yolo` only to explicitly grant full access.
+`--status` reports `executionBoundary`: managed `read_only`/`writable`, bypass, or external.
+A read-only task instruction does not turn a bypass Session into an enforced read-only Session.
+The legacy blocking `--resume <id>` selector retains its explicit `--yolo` requirement for bypass.
+
+An ordinary built-in AskUser returns `needs_input`, the questions and options, with exit code 0.
+The Host retains the waiting Turn. `--answer` accepts one string or `null` per question in a JSON
+array and continues the original tool call and Turn. It does not need `--resume`. A failed or
+uncertain answer submission is not automatically retried; read the current questions before
+choosing what to do next. Multiple pending batches are reported, and an answer is refused when
+the current batch cannot be selected uniquely.
+
+For Host-owned Session execution, `--wait` and `--status` return `needs_approval` with
+`approvalRequests` when a sandbox expansion is pending. Each entry contains its `kind`, requested
+`expansion`, and `justification`; internal request IDs stay inside the CLI. Review the scope, then:
+
+```sh
+maka run --session "$session" --approve --json
+maka run --session "$session" --deny --json
+```
+
+These are alternative decisions, not a sequence to execute. Both continue the original Turn and
+require exactly one pending sandbox request in the Session; none or multiple are rejected without
+submitting a decision. They do not need `--yolo` and cannot be combined with it. Approval applies
+the requested expansion through the Host's managed boundary; it does not change the Session to
+bypass. The expanded boundary is saved for subsequent execution in the Session. The flags
+target the request pending when that CLI invocation attaches. Do not automatically
+retry a decision after an uncertain result: read the current requests before deciding again.
+Ordinary blocking runs retain their existing sandbox-boundary denial behavior.
+
+For coordination, submit follow-ups without blocking, then service the Session's interactions:
+
+```sh
+maka run --session "$session" "Inspect the tests next" --no-wait
+maka run --session "$session" --wait --json
+```
+
+Worker admission receipts have `submission: "started"`, `"queued"`, or `"steering"`. Their `status`
+describes the currently observed Session, not completion of the submitted message. A queued receipt
+can include `needs_input` and the current Turn's questions, or `needs_approval` and its sandbox
+request. Answer that interaction before expecting the follow-up to execute. `queued` is the
+observed total of follow-up and steering messages; it is not an exact position. An admission
+receipt omits this count until its queue projection has caught up; use status for a fresh view. Status also reports
+`providerRetry` when the Host has published a retry, without guessing activity from silence.
+
+Assistant Inbox conversations instead return `submission: "inbox"`, without a worker result or queue count, even with `--wait`. This confirms Inbox admission only; inspect delegated work through its own worker Session.
+
+Queue is the default: each explicit queued message opens its own subsequent Turn. Steer uses the
+Host's existing current-Turn delivery and may carry over if it misses that Turn's last boundary.
+Submitting a queued prompt without `--no-wait` follows that message's execution (`--wait` is optional). It does not return
+questions from an earlier Turn blocking the queue; service those through a separate Session
+wait/answer call, or use the coordination workflow above. Steer with `--wait`
+observes the Turn selected on attachment; it does not guarantee consumption before that Turn ends. A standalone `--wait` fixes its
+target on attachment; when idle, it reads the latest result. `--result` never waits for an active
+Turn. Status and result queries do not prepare a model or start MCP. Interrupting Session
+observation with Ctrl+C leaves Host execution running; legacy blocking run cancellation retains
+its existing behavior. To stop Host execution explicitly, use `maka run --session "$session" --stop --json`.
+Stop targets the current Turn observed on attachment and needs no `--yolo`; a successful request
+exits 0. A successful explicit or implicit status query also exits 0 even if it reports a failed
+or cancelled Turn; result/wait retain the task's failure exit code. Stop and status are not
+assertions that the assigned task succeeded.
+
+Permission requests, MCP forms, and other interactions that this Session CLI cannot answer
+remain waiting in the Host. A wait reports the actual interaction kind and a recovery command: use an
+interactive client on the same Host, or stop the Turn explicitly. Detached admission does not
+promise that subsequent execution will need no interaction.
+
+`--session <id> --resume` instead requests safe continuation of
+stopped execution. An idle Session prompt can use `--max-steps`; a queued or steering message
+cannot attach a step cap to an existing Turn.
+
+Returning from client-owned MCP interaction and resuming it in another CLI process is not
+supported. Blocking MCP runs retain their existing behavior; `--no-wait` refuses to detach the
+MCP servers owned by that invocation. Graph interaction handoff is also deferred. Use a persistent
+CLI installation for local Session control execution and detached runs; temporary npx invocations
+may own the Host's lifetime and keep the legacy blocking path.
+
+For agent-to-agent CLI workflows, install the repository's `maka-cli` skill after it is available
+on your selected branch. It explains Session coordination; it does not install Maka itself:
+
+```sh
+npx skills add apache/maka --skill maka-cli
+```
+
 ## Upgrade
 
 Update within the nightly line by pinning the exact release (`--target` accepts `latest`,

package/README.zh-CN.md

@@ -108,6 +108,162 @@ maka run --help
 Maka 默认会在执行高权限工具操作前询问。`maka run --yolo` 会授予该任务完整的文件和网络
 权限，只应在你允许任务修改的环境中使用。
 
+
+主代理跨多次调用协调任务时，推荐使用 `--no-wait`，把返回的 `sessionId` 保存为 `$session`，
+准备处理交互或接收结果时再等待：
+
+```sh
+maka run "检查这个项目" --no-wait
+maka run --session "$session" --wait --json
+```
+
+`--no-wait` 本身就返回 JSON，无需再加 `--json`。普通 `maka run` 仍默认等待并输出最终答案，
+适合单次 shell 调用；需要结构化的阻塞结果时使用 `maka run "检查这个项目" --json`。
+
+根据目的选择下一次会话调用：
+
+```sh
+maka run --session "$session" --status
+maka run --session "$session" --recap --json
+maka run --session "$session" --answer '["是",null]' --json
+maka run --session "$session" "接下来检查测试" --no-wait
+maka run --session "$session" "重点检查参数解析" --steer --no-wait
+maka run --session "$session" --result --json
+maka run --session "$session" --resume --json
+```
+
+追加 prompt 默认排队，无需重复 `--queue`。重复相同参数或相同值可以接受；值冲突或无法
+同时执行的动作组合仍返回用法错误。执行命令默认等待，指定 `--no-wait` 才提交后返回；
+`--session <id> --wait` 则是不提交新工作，只观察当前 Turn。
+
+| 调用 | 默认 stdout | 退出码 0 的含义 |
+| --- | --- | --- |
+| 普通 run 或会话 wait/result | 最终答案文本；待处理交互返回 JSON | 仍须检查结果，待处理交互需要调用方决定 |
+| `--no-wait` | JSON 接收回执 | 提交被接收，不代表任务完成 |
+| `--status` / `--stop` | JSON | 状态查询或停止请求成功 |
+| 模型设置且无动作 | 带 `configuration` 的 JSON | 设置已保存，执行状态不变 |
+| `--recap` | 摘要文本 | 独立摘要生成成功 |
+
+在 wait、result、recap 上加 `--json` 可统一为结构化结果。`status` 表示生命周期，
+`finalOutput` 是任务答案，`recap` 是进展，`submission` 是接收回执。
+远程 Host 的每次调用都应携带相同的 `--host <profile-id>`。
+
+拒绝审批会将决定反馈给模型，同一个 Turn 可以在现有权限内继续或说明阻碍。工具失败本身不会
+覆盖 Runtime 的终态。wait 等到 Turn 结束或下一次交互时返回；`--deny --no-wait` 只提交决定。
+错误默认以简短文本输出到 stderr；显式 `--json` 才向 stdout 输出结构化错误，保留原始信息和
+Host 已提供的 `code`。操作失败不代表 Session 已失败。
+
+`--session` 继承已保存的设置，并允许持久更新模型、连接和 thinking。工作目录、timeout 和
+Graph 覆盖仍需使用旧选择器 `--resume <id>`；与 `--session` 组合时会在连接前拒绝。
+
+会话模型切换会持久保存，再由显式参数决定续跑或发送新消息：
+
+```sh
+maka run --session "$session" --model "other-model"
+maka run --session "$session" --model "other-model" --resume --json
+maka run --session "$session" --model "other-model" "继续审查" --no-wait
+```
+
+这三条是不同用法。第一条只保存设置，返回带有 `connection`、`model`、`thinking` 的 JSON
+`configuration` 回执，不会启动或恢复执行。加 `--connection <slug>` 可以选择另一个已保存的
+账号。原生模型或连接实际改变时会将 thinking 重置为模型默认值，除非显式指定 `--thinking`；选择相同目标会保留已保存的级别。
+与 Desktop 切换模型一样，新设置对后续调用持续生效；全局默认值和权限不变。
+
+设置先保存，再执行显式动作；即使后续动作失败，已保存的设置仍然保留。CLI 不判断会话因何
+停止，也不自动选择模型。Host 返回忙碌等错误时会直接返回，不隐式停止 Turn 或重发旧 prompt。
+resume 使用 Host 已有的安全续跑操作；提供新 prompt 则是新输入。
+
+`--recap` 复用 TUI `/recap` 的独立模型调用，用一句话总结会话已记录的进展。
+它可以在 Turn 运行中使用，不会 steering、排队、停止当前 Turn，也不等 Turn 完成。
+不加 `--json` 时直接输出摘要；加上后返回 `sessionId`、当前 `status` 和独立的 `recap` 字段，例如：
+
+```json
+{"sessionId":"…","status":"running","queued":0,"recap":"你要求审查代码；参数解析已检查，测试验证尚未完成。"}
+```
+
+摘要基于生成开始时读取的历史快照，无法看到尚未记录的工作或尚未返回的工具结果；
+生成期间任务可能继续推进。每次调用都会重新请求模型并产生模型用量；只查状态继续用
+`--status`。摘要不是最终答案（`--result`），也不是正在运行的代理直接回复的消息。
+`--status --recap` 可组合返回包含状态、执行边界和进展的 JSON；recap 不能与其他会话动作、
+prompt 或等待参数组合使用。
+摘要生成成功返回退出码 0，即使会话最近的 Turn 失败；生成失败返回退出码 1。
+Ctrl+C 取消 CLI 观察，不会停止 Host 中的执行。
+
+会话控制（`--session`）与继承模型、工作目录一样，继承 Host 已保存的权限。续跑、queue、
+steer 和回答问题不需要重复 `--yolo`，也不会改变执行边界；只有要显式授予完整权限时才加
+`--yolo`。`--status` 会报告 `executionBoundary`：受管 `read_only`/`writable`、bypass 或 external。
+只读任务描述不等于强制只读沙箱，继承 bypass 的会话依然是完整权限。旧阻塞选择器
+`--resume <id>` 对 bypass 会话仍保留显式 `--yolo` 要求。
+
+内置 AskUser 是普通阻塞运行的例外：它会输出问题 JSON 而不是最终答案。脚本应使用 `--json`
+并检查 `status`，不能仅凭退出码 0 判断任务完成。普通内置 AskUser 会以退出码 0 返回
+`needs_input`、问题及选项，Host 保留等待中的 Turn。
+`--answer` 使用 JSON 数组，为每个问题提供一个字符串或 `null`，回答后原工具调用和原 Turn
+继续执行，无需 `--resume`。答案提交失败或结果不确定时不会自动重发；先重新读取当前问题，
+再决定下一步。存在多批待回答问题时会如实报告，无法唯一选择当前批次时拒绝提交答案。
+
+对于 Host 持有的会话执行，`--wait` 和 `--status` 遇到沙箱扩展请求会返回 `needs_approval`，
+并在 `approvalRequests` 中列出 `kind`、扩展范围 `expansion` 和理由 `justification`；内部请求
+ID 不暴露。查看范围后，选择批准或拒绝：
+
+```sh
+maka run --session "$session" --approve --json
+maka run --session "$session" --deny --json
+```
+
+上面两条命令是二选一，不能顺序执行。它们都继续原 Turn，要求会话中恰好有一个待处理沙箱
+请求；没有请求或存在多个请求时拒绝提交决定。它们无需 `--yolo`，也不能与之混用。批准通过
+Host 扩展受管沙箱边界，并保存供该会话后续执行使用，不会将会话切换为 bypass。命令针对
+本次连接时待处理的请求；结果
+不确定时不要自动重试，应先读取当前请求，再决定下一步。普通阻塞运行仍保留原有的沙箱
+边界自动拒绝行为。
+
+协调线程推荐先后台入队，再处理会话交互：
+
+```sh
+maka run --session "$session" "接下来检查测试" --no-wait
+maka run --session "$session" --wait --json
+```
+
+工作会话提交回执中的 `submission` 为 `started`、`queued` 或 `steering`，只表示接收方式；`status`
+描述当前观察到的会话，不表示刚提交的消息已完成。排队回执可同时带有 `needs_input` 和当前
+Turn 的问题，或 `needs_approval` 和沙箱请求；先回答交互，后续消息才能执行。`queued` 是观察到
+的后续消息与 steer 总数，不是精确排队位置；回执快照未追上接收结果时省略该计数，
+可用 status 读取新状态。Host 发布重试状态时，status 还会带上
+`providerRetry`，不会根据静默时间猜测活动。
+
+助手 Inbox 会话返回 `submission: "inbox"`，不包含工作 Turn 的结果或队列计数，即使指定 `--wait` 也立即返回接收回执。这仅确认消息已进入 Inbox；委派工作的结果需通过对应工作会话查询。
+
+默认使用 queue，每条显式排队消息对应一个独立后续 Turn。steer 沿用 Host 当前 Turn 的追加
+机制，错过最后处理边界时可能进入后续 Turn。提交排队消息且未指定 `--no-wait` 时，调用跟踪该消息对应的执行（可显式加 `--wait`），不会返回前一 Turn 阻塞队列的问题；
+这类问题要通过独立的会话 wait/answer 处理，或采用上面的协调流程。steer 配合 `--wait` 观察连接时选定的
+Turn，不保证追加消息一定在该 Turn 结束前消费。
+单独 `--wait` 在连接时固定目标，空闲时读取最近结果。`--result` 不等待仍在执行的 Turn。
+状态和结果查询不准备模型、不启动 MCP。用 Ctrl+C 退出会话观察会让 Host 继续执行；旧阻塞
+run 的取消行为保持原样。显式停止 Host 执行可用 `maka run --session "$session" --stop --json`。
+停止操作针对连接时观察到的当前 Turn，不需要 `--yolo`，请求成功退出 0。显式或隐式的 status
+查询成功时也退出 0，即使它报告的 Turn 已失败或取消；result/wait 仍以任务失败返回非零退出码。
+停止或查询成功不代表分配的任务已完成。
+
+权限请求、MCP 表单等会话 CLI 无法回答的交互会留在 Host 中等待。wait 会报告实际交互类型和恢复
+命令：使用同一 Host 的交互式客户端处理，或显式停止该 Turn。后台提交成功不保证后续执行
+不会请求交互。
+
+`--session <id> --resume` 则表示
+安全恢复已停止的执行。空闲会话的新 prompt 可以使用 `--max-steps`，排队或 steer 消息不能
+给已有 Turn 附加步数上限。
+
+当前不支持客户端 MCP 交互返回后由另一个 CLI 进程续接。阻塞 MCP 运行保留现有行为；
+`--no-wait` 会拒绝脱离本次调用持有的 MCP servers。Graph 交互交接也暂缓。执行本地会话控制
+及后台提交时使用持久 CLI 安装；临时 npx 调用可能持有 Host 生命周期，继续使用旧阻塞路径。
+
+仓库的 `maka-cli` skill 提供主/子代理 CLI 协作指引；待其进入所选分支后可安装。
+skill 本身不安装 Maka CLI：
+
+```sh
+npx skills add apache/maka --skill maka-cli
+```
+
 ## 升级
 
 在 nightly 线内更新时固定精确版本（`--target` 只接受 `latest`、`next` 或精确的 Maka

package/THIRD_PARTY_NOTICES.txt

@@ -752,9 +752,9 @@ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
 ================================================================================
 
-Package: @modelcontextprotocol/client@2.2.0
-Declared license: MIT
-Selected license: MIT
+Package: @modelcontextprotocol/client@2.3.0
+Declared license: Apache-2.0
+Selected license: Apache-2.0
 Repository: git+https://github.com/modelcontextprotocol/typescript-sdk.git
 
 --- LICENSE ---
@@ -977,9 +977,9 @@ the full license text.
 
 ================================================================================
 
-Package: @modelcontextprotocol/core@2.2.0
-Declared license: MIT
-Selected license: MIT
+Package: @modelcontextprotocol/core@2.3.0
+Declared license: Apache-2.0
+Selected license: Apache-2.0
 Repository: git+https://github.com/modelcontextprotocol/typescript-sdk.git
 
 --- LICENSE ---
@@ -3970,7 +3970,7 @@ SOFTWARE.
 
 ================================================================================
 
-Package: fast-uri@3.1.8
+Package: fast-uri@3.1.9
 Declared license: BSD-3-Clause
 Selected license: BSD-3-Clause
 Repository: git+https://github.com/fastify/fast-uri.git
@@ -4027,7 +4027,7 @@ THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLI
 
 ================================================================================
 
-Package: follow-redirects@1.16.0
+Package: follow-redirects@1.16.1
 Declared license: MIT
 Selected license: MIT
 Repository: git+ssh://git@github.com/follow-redirects/follow-redirects.git
@@ -4781,7 +4781,7 @@ OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
 ================================================================================
 
-Package: ip-address@10.7.2
+Package: ip-address@10.7.4
 Declared license: MIT
 Selected license: MIT
 Repository: https://github.com/beaugunderson/ip-address.git
@@ -7444,7 +7444,7 @@ PERFORMANCE OF THIS SOFTWARE.
 
 ================================================================================
 
-Package: undici@7.30.0
+Package: undici@7.30.1
 Declared license: MIT
 Selected license: MIT
 Repository: git+https://github.com/nodejs/undici.git
@@ -7474,7 +7474,7 @@ SOFTWARE.
 
 ================================================================================
 
-Package: undici@8.11.2
+Package: undici@8.12.3
 Declared license: MIT
 Selected license: MIT
 Repository: git+https://github.com/nodejs/undici.git
@@ -7934,7 +7934,7 @@ ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
 
 ================================================================================
 
-Package: zod@4.6.5
+Package: zod@4.6.7
 Declared license: MIT
 Selected license: MIT
 Repository: git+https://github.com/colinhacks/zod.git

package/dist/cli-core.js

@@ -316,6 +316,7 @@ export async function runMakaCli(argv = process.argv.slice(2), options = RELEASE
                 ...(command.repairRootAfterRemount ? { repairRootAfterRemount: true } : {}),
                 updateExisting: command.updateExisting,
                 reuseExistingEnvironment: command.reuseExistingEnvironment,
+                ...(command.expectedExisting ? { expectedExisting: command.expectedExisting } : {}),
                 allowInterruptActiveTasks: command.allowInterruptActiveTasks,
                 ...(command.rootPath ? { rootPath: command.rootPath } : {}),
                 ...(command.projectDirectoryRoots

package/dist/pi-tui-runner.js

@@ -3786,7 +3786,7 @@ export async function runMakaPiTui(input) {
             selectedIndex: items.findIndex((item) => item.value === cursorValue),
         });
     };
-    const showGoalSummary = () => {
+    const showGoalSummary = async () => {
         // Read the live projection at request time: /goal is the one place the
         // user asks for the state *right now*.
         if (!input.driver.getGoal) {
@@ -3798,11 +3798,33 @@ export async function runMakaPiTui(input) {
             requestRender();
             return;
         }
+        const sessionId = input.driver.getSessionId();
+        const waits = input.driver.queryEventWaits
+            ? await input.driver.queryEventWaits()
+            : (input.driver.getEventWaits?.() ?? []);
+        if (input.driver.getSessionId() !== sessionId)
+            return;
         const goal = input.driver.getGoal();
         state.entries.push({
             kind: 'notice',
             level: 'info',
-            text: goal ? goalSummaryLines(goal, Date.now()).join('\n') : 'No goal set.',
+            text: [
+                goal ? goalSummaryLines(goal, Date.now()).join('\n') : 'No goal set.',
+                ...waits.map((wait) => {
+                    const copy = TUI_COPY_RESOURCES.goal[locale];
+                    const observations = {
+                        checking: copy.waitChecking,
+                        pending: copy.waitPending,
+                        unknown: copy.waitUnknown,
+                    };
+                    const label = wait.phase === 'result_ready'
+                        ? copy.waitResultReady
+                        : wait.observation === null
+                            ? copy.waitObserving
+                            : observations[wait.observation];
+                    return `${wait.waitId}: ${label} · ${wait.resource.resourceId}${wait.deadlineAt === null ? '' : ` · ${Math.abs(wait.deadlineAt) <= 8.64e15 ? new Date(wait.deadlineAt).toISOString() : String(wait.deadlineAt)}`}`;
+                }),
+            ].join('\n'),
         });
         requestRender();
     };
@@ -4058,16 +4080,61 @@ export async function runMakaPiTui(input) {
                     // Read-only, so no runControl busy gate: an autonomous loop keeps
                     // the session busy almost by definition, and that is exactly when
                     // the user wants to inspect it.
-                    showGoalSummary();
+                    void showGoalSummary().catch((error) => {
+                        state.entries.push({ kind: 'notice', level: 'error', text: String(error) });
+                        requestRender();
+                    });
                     return;
                 }
                 const action = parts[1];
+                if ((parts.length === 2 || parts.length === 3) &&
+                    action === 'cancel-wait' &&
+                    input.driver.cancelSessionEventWait) {
+                    void (async () => {
+                        const waits = input.driver.queryEventWaits
+                            ? await input.driver.queryEventWaits()
+                            : (input.driver.getEventWaits?.() ?? []);
+                        const wait = parts[2]
+                            ? waits.find((candidate) => candidate.waitId === parts[2])
+                            : waits.length === 1
+                                ? waits[0]
+                                : undefined;
+                        if (!wait) {
+                            state.entries.push({
+                                kind: 'notice',
+                                level: 'info',
+                                text: waits.length > 1
+                                    ? `${TUI_COPY_RESOURCES.goal[locale].chooseWait}\n${waits.map((candidate) => `${candidate.waitId} · ${candidate.resource.resourceId}`).join('\n')}`
+                                    : TUI_COPY_RESOURCES.goal[locale].noCancellableWait,
+                            });
+                        }
+                        else {
+                            const result = await input.driver.cancelSessionEventWait({
+                                sessionId: wait.sessionId,
+                                waitId: wait.waitId,
+                                consumer: wait.consumer,
+                            });
+                            state.entries.push({
+                                kind: 'notice',
+                                level: 'info',
+                                text: result.outcome === 'cancelled'
+                                    ? TUI_COPY_RESOURCES.goal[locale].sessionWaitCancelled
+                                    : TUI_COPY_RESOURCES.goal[locale].waitChanged,
+                            });
+                        }
+                        requestRender();
+                    })().catch((error) => {
+                        state.entries.push({ kind: 'notice', level: 'error', text: String(error) });
+                        requestRender();
+                    });
+                    return;
+                }
                 if (parts.length !== 2 ||
                     (action !== 'pause' && action !== 'resume' && action !== 'clear')) {
                     state.entries.push({
                         kind: 'notice',
                         level: 'error',
-                        text: 'Usage: /goal [pause|resume|clear]',
+                        text: TUI_COPY_RESOURCES.goal[locale].usage,
                     });
                     requestRender();
                     return;