docs(rules): one Copilot review per PR, another only while high findings remain

.claude/skills/open-pr/SKILL.md

@@ -189,13 +189,16 @@ re-reviewed. Two consequences for this loop. A FIX push starts no new
 Copilot run, so a `copilot-*` check on the new head SHA is
 legitimately ABSENT; waiting for one that will never come is the
 failure mode to avoid — see §3e for what to require instead. And a
-fresh review is requested only after a SUBSTANTIVE rework (new
-behaviour, a reworked mechanism), never after every push: each
-request re-reads the whole diff and surfaces "previously missed"
-findings in files the push never touched, which draws another push
-(kurrentschrift#406 collected ~15 requests in a day over a one-line
-docstring fix). Stop once a round yields no new inline comments but
-only carried-over items.
+further review is requested only while the previous round found
+something high or critical — a real bug, a security or correctness
+issue, a wrong mechanism — whose fix you pushed; a round with only
+low and medium findings gets its fixes pushed and its threads
+resolved, and that is the end (owner, 2026-10-23: each round costs
+premium requests, and a docs PR had collected three over wording).
+Each request re-reads the whole diff and
+surfaces "previously missed" findings in files the push never
+touched, which draws another push (kurrentschrift#406 collected ~15
+requests in a day over a one-line docstring fix).
 
 Fetch all three comment surfaces — they carry different content:
 
@@ -286,7 +289,8 @@ stale wait loop with the bracket trick (`pkill -f "x[.]y"`), or
 `.claude/skills/open-pr/scripts/` cover the loops, run from the
 repository root: `watch-checks.sh <num>` (§3a),
 `watch-copilot.sh <num> [count]` (§3b–c: waits for `count` Copilot
-reviews, 2 for the extra round after fixes, then prints the latest
+reviews, 2 only when a high finding was fixed and another round
+was requested, then prints the latest
 summary, every inline comment and the unresolved thread ids) and
 `watch-deploys.sh <merge-sha>` (§4).
 
@@ -392,11 +396,12 @@ worktree stay; deleting those still needs the owner's OK.
   thread while it stays unresolved; outdated threads still count
   against review-clean — resolve them explicitly.
 - **A fix push no longer starts a review round.** `review_on_push` is
-  `false` since 2026-09-03 (§3b), so only an explicit — and
-  substantive — re-request opens another one. When a round does run,
-  new threads on the changed lines are the loop working, not noise —
-  but don't chase cosmetic nits past a couple of rounds; surface
-  stalemates to the user.
+  `false` since 2026-09-03 (§3b), so only an explicit re-request opens
+  another one, and that only while the previous round had high or
+  critical findings whose fixes you pushed (§3b). A round with only
+  low and medium findings ends the loop: push the fixes, resolve the
+  threads, never chase nits into another round; surface a stalemate
+  to the user instead.
 - **Stacked PRs die when their base squash-merges.** Don't stack; if
   work depends on an unmerged PR, wait for its merge (or do the work
   and rebase before opening).

.claude/skills/open-pr/scripts/watch-copilot.sh

@@ -7,7 +7,7 @@
 # than a PR collects.
 #
 # Usage: watch-copilot.sh <pr> [count, default 1]
-#   count 2 waits for the one extra round requested after fixing the first round's findings.
+#   count 2 waits for a further round, requested only after a high or critical finding was fixed.
 set -uo pipefail
 PR=${1:?usage: watch-copilot.sh <pr> [count]}
 WANT=${2:-1}

CLAUDE.md

@@ -42,7 +42,7 @@ After opening a PR, the work is **not** complete. Stay with the PR until both th
 
 1. **Watch the pipeline.** Poll `gh pr checks <num>` (and Cloud Build for triggered deploys) until every required check has finished. Use a background bash poll so other work can continue. Default: poll every 20 s, up to ~10 min per check.
 2. **Fix CI failures.** If any check fails, read the relevant log (`gh run view --log-failed`, `gcloud builds log <id>`), push a fix commit to the same branch, then keep watching. Repeat until green.
-3. **Wait for the Copilot PR Reviewer bot — ONE review per PR, not one per push.** Review-on-push is off in the "Automated Copilot Code Review" ruleset (owner, 2026-09-03): the bot runs once when the PR opens or leaves draft, and a push triggers nothing. Don't re-request a review after every push. Each request is a full re-read of the whole diff, and the bot then surfaces "previously missed" findings in files the push never touched — which draws another push, which draws another request (the sibling repo's [kurrentschrift#406](https://github.com/MarkusNeusinger/kurrentschrift/pull/406) collected ~15 requests in a day over a one-line docstring fix). Request a fresh review explicitly only after a SUBSTANTIVE rework (new behaviour, a reworked mechanism), and stop once a round yields no new inline comments but only carried-over items: the field is grazed. A PR that is green with no open threads needs no further round — say so and let the owner merge. The first review typically lands within ~2 min of PR open. Fetch with `gh pr view <num> --comments`, plus the three GitHub APIs that surface different comment types — `gh api` resolves `{owner}/{repo}` from the current git remote so these are copy/paste-portable:
+3. **Wait for the Copilot PR Reviewer bot — ONE review per PR, not one per push.** Review-on-push is off in the "Automated Copilot Code Review" ruleset (owner, 2026-09-03): the bot runs once when the PR opens or leaves draft, and a push triggers nothing. Don't re-request a review after every push. Each request is a full re-read of the whole diff, and the bot then surfaces "previously missed" findings in files the push never touched — which draws another push, which draws another request (the sibling repo's [kurrentschrift#406](https://github.com/MarkusNeusinger/kurrentschrift/pull/406) collected ~15 requests in a day over a one-line docstring fix). Request another review only while the previous round found something high or critical — a real bug, a security or correctness issue, a wrong mechanism — whose fix you pushed; a round with only low and medium findings gets its fixes pushed and its threads resolved, and that is the end (owner, 2026-10-23: every round costs premium requests; a docs PR had collected three over wording). A PR that is green with no open threads needs no further round — say so and let the owner merge. The first review typically lands within ~2 min of PR open. Fetch with `gh pr view <num> --comments`, plus the three GitHub APIs that surface different comment types — `gh api` resolves `{owner}/{repo}` from the current git remote so these are copy/paste-portable:
    - `gh api repos/{owner}/{repo}/pulls/<num>/reviews` — top-level review summaries (Copilot's overall comment lives here)
    - `gh api repos/{owner}/{repo}/pulls/<num>/comments` — inline review comments tied to file/line
    - `gh api repos/{owner}/{repo}/issues/<num>/comments` — generic PR conversation comments (codecov, deployment bots, humans)

agentic/commands/pull_request.md

@@ -65,7 +65,8 @@ EOF
 8. **The Copilot review runs ONCE**, when the PR opens (or leaves draft) — review-on-push is off in
    the "Automated Copilot Code Review" ruleset (owner, 2026-09-03), so step 7's follow-up commit and
    every later push trigger nothing. Do not re-request a review per push; the rule and its one
-   exception (a substantive rework) are in `CLAUDE.md` § "PR Follow-Through", step 3.
+   exception (another round only while the previous one had high or critical findings) are in
+   `CLAUDE.md` § "PR Follow-Through", step 3.
 
 ## Report
 

changelog.d/copilot-one-round.md

@@ -0,0 +1,9 @@
+### Changed
+
+- **Copilot reviews a PR once; another round only while the previous one
+  had high or critical findings.** `CLAUDE.md`, the `open-pr` skill and
+  `/pull_request` replace "stop once a round yields nothing new", which let
+  a round with any findings draw the next one, with the owner's rule of
+  2026-10-23: a round with only low and medium findings gets its fixes
+  pushed and its threads resolved, and that is the end, because every round
+  re-reads the whole diff and costs premium requests. (#12250)