feat(skills,pipelines): ask the user at the two points being wrong is cheap, and stop paying for reviews the diff does not need

The `default` 8-node pipeline walked past its own intake: the node declared
`agenthub stage intake`, a stage report with no work attached, so "intake" was
a label the run passed through without ever asking a question. It now declares
`/superpowers:brainstorming` (and detects it — that rule moves off `plan`),
and both driving skills make the brainstorm mandatory, asked with
AskUserQuestion and waited on rather than answered on the user's behalf.

Three stops replace three silent slides:

- After the plan is written and adversarially challenged, it goes to the USER
  for approval. Implementation does not start without it, and work that turns
  out to need a different approach comes back for a fresh one.
- Review picks its reviewers from `git diff --name-only`: Claude Code's own
  `code-review` always, a domain reviewer only for a layer the diff actually
  touched, and — new — only once the change is big enough to have seams. Three
  agents re-reading twenty lines buys one opinion at three times the tokens, on
  every fix-loop lap. Review is also a loop now: it ends on a clean pass over
  the current diff, not on one pass having happened.
- The gate asks the user to accept before anything is claimed done (and in the
  lane pipeline, before the PR is published — publishing is outward-facing).
  A NO-GO re-enters by SIZE: a defect goes back to implement, a wrong approach
  back to plan. Nothing carries a NO-GO forward with a note about it.

Also here, from a real stall seen on a lane: `ship-feature` had no waiting rule
at all, and a session at the gate invented one with `ScheduleWakeup` — a
/loop-only primitive that paces nothing here and is refused outright when
`auto` mode cannot reach its safety classifier. The rule the lane skill already
had now exists in both.

Cost, measured in what each agent is actually for: the suite run leaves its
agent (nothing is judged, so nothing is delegated — an agent there buys a
second model's tokens to read output the session can read for free), and
`backend-reviewer` drops from opus to sonnet like its frontend twin; both read
a scoped diff against a fixed checklist, and the senior gate above them is
still opus.

Skill bugs found in the same sweep: doc-map told agents not to recreate the
removed translations and then, three lines later, to mirror every README change
into them; the docs mapping still carried `+VN+CN` and the punctuation left by
their removal; Procedure jumped 2 → 6; Verify required `scripts/doc-coverage.sh`,
which does not exist in this repo.
This commit is contained in:
2026-09-18 10:21:03 +07:00
parent 4227a97ab1
commit 26b0fe53c7
13 changed files with 180 additions and 80 deletions
+2 -2
View File
@@ -1123,8 +1123,8 @@ a `detect` array on each node. The built-in `default.json`
| Node | Detect rules |
|---|---|
| `intake` | `Agent` matching subagent `Explore` — a lane sitting at nothing else yet, but already exploring, shows *something* instead of idle |
| `plan` | `Skill` matching `brainstorming\|writing-plans`; `Write` matching `docs/.*plan.*\.md`; `Agent` matching `Plan\|Software Architect\|Backend Architect\|UX Architect`; `TaskCreate` (any) |
| `intake` | `Agent` matching subagent `Explore` — a lane sitting at nothing else yet, but already exploring, shows *something* instead of idle; `Skill` matching `brainstorming`, which is intake's own declared command |
| `plan` | `Skill` matching `writing-plans`; `Write` matching `docs/.*plan.*\.md`; `Agent` matching `Plan\|Software Architect\|Backend Architect\|UX Architect`; `TaskCreate` (any) |
| `implement` | `Skill` matching `executing-plans\|subagent-driven-development\|test-driven-development`; `Edit` (any); `Write` to any path NOT under a `docs/` directory (`^(?!.*(?:^\|/)docs/)`) |
| `tests` | `Bash` matching `\b(npm (run )?test\|pytest\|vitest\|jest\|go test\|cargo test\|node\s+--test\|rspec\|phpunit\|dotnet test\|mvn test\|gradle test)\b`; `Agent` matching `Test Results Analyzer\|API Tester\|Performance Benchmarker` |
| `review` | `Skill` matching `code-review\|requesting-code-review\|receiving-code-review`; `Bash` matching `\bgh\b[^;&\|]*\bpr diff\b`; `Agent` matching a name containing `reviewer`/`Reviewer` (excluding `senior-gate-reviewer`, a lookbehind), `Evidence Collector`, or `Accessibility Auditor` |
+2 -2
View File
@@ -212,8 +212,8 @@ any of them is ever deleted:
| The directory `agenthub.service` names (`unitRoot`, **foreign unit included**) | A service is executing from it. Not ours to rewrite is still not ours to delete. |
| Any directory named in the user's `settings.json` hooks | A Claude Code process that has not been restarted is still executing from there — this is the same directory the section above describes, seen from the other side. Deleting it would break that session's hooks mid-flight. |
Plus the newest `KEEP_RETIRED_ROOTS` (2) of whatever is left, as a rollback
cushion. Outside `~/.claude/plugins/cache/` the pruner does nothing at all — for
Plus the newest `KEEP_RETIRED_ROOTS` (1) of whatever is left, as a rollback
cushion — one build deep, because a full copy of this repo is ~130 MB. Outside `~/.claude/plugins/cache/` the pruner does nothing at all — for
a git checkout the parent directory holds the user's own work. Best effort
throughout: a directory that will not delete is left alone, never an error.