[linkerlin]

Summary

未完成 run 上，软缺口（soft_continue=true）的 swarm_doctor 已经指向 swarm_solidify。只在 tick_again 严格为 false 时才停、把缺席当成再 tick、同时不看 next_action 的宿主，冒烟之后仍会再 tick。这次在这条信封上写 tick_again=false。soft_continue 仍为 true，表示缺口不是硬停。不写 gate_soft_continue，不写 session_not_failed。本调用不结算这条 run，不写 Parent baseline，不执行 evolver doctor --fix，也不放宽固化门。第一次 tick 重定向和保险丝本来就有 tick_again=false，这次不改。

On an unfinished run, a soft-continue swarm_doctor already named swarm_solidify. A host that stops only when tick_again is strict false, and treats an omitted key as another tick, still ticked after the smoke because the key was absent. This writes tick_again=false on that envelope. soft_continue stays true: the gap is not a hard stop. gate_soft_continue and session_not_failed stay absent. The call does not settle the run, write a Parent baseline, run evolver doctor --fix, or relax the solidify gate. The first unfinished-run tick redirect and the fuse already set tick_again false; those envelopes are unchanged.

What changed

• 仅当存在未完成 run、soft_continue=true、且没有硬 blockers 时，swarm_doctor 增加 tick_again=false。next_action 仍是 swarm_solidify，preserve 三元组不动。hint 写明 tick_again=false，并写明 fixes 里的 evolver doctor --fix 本工具不执行、固化前不要跑。
• soft_continue 仍为 true。不写 gate_soft_continue=false（会重开 DEBUG #67）。不写 session_not_failed（那是 tick 重定向的旗标，不是 doctor 的健康位）。
• 没有未完成 run 时 doctor 不写 tick_again。硬阻断（例如不是 git 仓库，ok=false）也不写。
• 工具描述与返回体同一口径。DEBUG #61 的宿主目录缓存没有做假修复。
• 不改第一次 tick 重定向（仍是 ok=true、retryable=false、tick_again=false、没有 error、没有 loop_fuse、软缺口上 gate_soft_continue 缺席）。不改保险丝（ok=false、loop_fuse=pending_redirect_loop、tick_again=false）。

How to test

1. uv run pytest tests/test_mcp_host_usability_pending_smoke.py tests/test_mcp_server.py tests/test_mcp_protocol.py tests/test_mcp_host_usability_p0.py tests/test_mcp_host_usability_p1.py tests/test_mcp_host_usability_p2.py tests/test_mcp_host_usability_polish.py tests/test_swarm.py tests/test_swarm_beginner_guidance.py tests/test_swarm_instrument_soft_copy.py tests/test_swarm_status_boot.py -q --tb=line
   • Expected: 163 passed.
   • test_pending_doctor_on_a_soft_gap_names_solidify fails on unmodified master (KeyError: tick_again). The live doctor tool description also fails there until it names tick_again=false. The hard-blocker case stays ok=false with no tick_again. Idle doctor (no unfinished run) does not gain the key.
2. uv run ruff check src tests && uv run ruff format --check src tests && uv run mypy src
   • Expected: ruff clean (746 files already formatted), mypy Success: no issues found in 353 source files.

Risk

Low. Only the soft-continue unfinished-run doctor envelope gains tick_again=false. ok, soft_continue, and loop_ready stay as the open-box report computed them. Hard failures and the idle doctor do not gain the key. Solidify gate, skip_validation, Parent baseline, and val seal are unchanged. The fuse still returns ok=false.

Harness/evaluator governance

Upstream governance surface: MCP swarm_doctor return while a run is unfinished (soft-continue only). Not solidify, selector, mutation, or the frozen gate.
Downstream EvoX impact: N/A -- existing read-only tool return, no bridge or contract change.
Rollout-local scope: host-visible doctor smoke only; boot, status, and the tick redirect envelope are unchanged.
Promotion boundary: N/A -- no proposal/rollout/default promotion.
Evaluator mismatch sets: N/A
Non-regression evidence: tests/test_mcp_host_usability_pending_smoke.py pins tick_again=false beside soft_continue=true, the preserve triad, absent gate_soft_continue / session_not_failed, the hint that this tool does not execute evolver doctor --fix, and the hard-blocker / idle paths (no tick_again). Tick redirects already asserted tick_again false; that assertion now also uses the same host model.
Fix-severity review: low
Owner approval: N/A
Security boundary: no new data, tool, host, network, or secret surface. Doctor still calls doctor(fix=False).
Rollback: revert this commit.
Live promotion: no
Autonomous evaluator self-editing: no

Self-check

• No new runtime dependencies without a clear justification in "What changed".
• Tests added or updated to cover the new behavior; gates pass locally: uv run pytest (modules listed above, 163 passed), uv run ruff check src tests, uv run ruff format --check src tests, uv run mypy src (353 source files).

Residual host friction

• 只读 soft_continue=true、完全不看 tick_again / next_action / ok / retryable 的宿主，doctor 冒烟之后仍可能再 tick。没有把 soft_continue 写成 false。
• 把缺席的 gate_soft_continue 当成继续、且不看 tick_again 的 tick 重定向（DEBUG #68 余项 1）仍可能再 tick。第一次重定向和保险丝已经写了 tick_again=false；看这个键的宿主会停。
• 把保险丝 ok=false 当成整段会话失败、且不读 session_not_failed 与 loop_fuse 的宿主，仍会在第二次 tick 上停掉会话。ok 留着 false，这次不改。
• swarm_status / swarm_boot 在未完成 run 上仍是 gate_soft_continue=true，不带 tick_again。只认这个 true 的宿主仍会 tick。status 的 next_action 仍是 swarm_solidify。
• fixes 仍列出 evolver doctor --fix。本工具不执行它。hint 写明固化前不要跑。
• 宿主 tools/list 缓存滞后（DEBUG #61）没有做假修复。

  

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota. Try requesting a review at a lower effort level.