Skip to content

Commit cfb6b56

Browse files
committed
chore(wheelhouse): cascade template@cf3f2d2 (followup)
1 parent a5cd45a commit cfb6b56

3 files changed

Lines changed: 76 additions & 0 deletions

File tree

.gitattributes

Lines changed: 11 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -135,6 +135,11 @@
135135
.claude/hooks/file-size-reminder/package.json linguist-generated=true
136136
.claude/hooks/file-size-reminder/test/index.test.mts linguist-generated=true
137137
.claude/hooks/file-size-reminder/tsconfig.json linguist-generated=true
138+
.claude/hooks/follow-direct-imperative-reminder/README.md linguist-generated=true
139+
.claude/hooks/follow-direct-imperative-reminder/index.mts linguist-generated=true
140+
.claude/hooks/follow-direct-imperative-reminder/package.json linguist-generated=true
141+
.claude/hooks/follow-direct-imperative-reminder/test/index.test.mts linguist-generated=true
142+
.claude/hooks/follow-direct-imperative-reminder/tsconfig.json linguist-generated=true
138143
.claude/hooks/gitmodules-comment-guard/README.md linguist-generated=true
139144
.claude/hooks/gitmodules-comment-guard/index.mts linguist-generated=true
140145
.claude/hooks/gitmodules-comment-guard/package.json linguist-generated=true
@@ -461,6 +466,10 @@
461466
.claude/skills/handing-off/SKILL.md linguist-generated=true
462467
.claude/skills/locking-down-programmatic-claude/SKILL.md linguist-generated=true
463468
.claude/skills/plug-leaking-promise-race/SKILL.md linguist-generated=true
469+
.claude/skills/prose/SKILL.md linguist-generated=true
470+
.claude/skills/prose/references/examples.md linguist-generated=true
471+
.claude/skills/prose/references/phrases.md linguist-generated=true
472+
.claude/skills/prose/references/structures.md linguist-generated=true
464473
.claude/skills/refreshing-history/SKILL.md linguist-generated=true
465474
.claude/skills/refreshing-history/run.mts linguist-generated=true
466475
.claude/skills/reviewing-code/SKILL.md linguist-generated=true
@@ -529,6 +538,7 @@
529538
.config/oxlint-plugin/rules/prefer-function-declaration.mts linguist-generated=true
530539
.config/oxlint-plugin/rules/prefer-node-builtin-imports.mts linguist-generated=true
531540
.config/oxlint-plugin/rules/prefer-node-modules-dot-cache.mts linguist-generated=true
541+
.config/oxlint-plugin/rules/prefer-non-capturing-group.mts linguist-generated=true
532542
.config/oxlint-plugin/rules/prefer-safe-delete.mts linguist-generated=true
533543
.config/oxlint-plugin/rules/prefer-separate-type-import.mts linguist-generated=true
534544
.config/oxlint-plugin/rules/prefer-spawn-over-execsync.mts linguist-generated=true
@@ -573,6 +583,7 @@
573583
.config/oxlint-plugin/test/prefer-function-declaration.test.mts linguist-generated=true
574584
.config/oxlint-plugin/test/prefer-node-builtin-imports.test.mts linguist-generated=true
575585
.config/oxlint-plugin/test/prefer-node-modules-dot-cache.test.mts linguist-generated=true
586+
.config/oxlint-plugin/test/prefer-non-capturing-group.test.mts linguist-generated=true
576587
.config/oxlint-plugin/test/prefer-safe-delete.test.mts linguist-generated=true
577588
.config/oxlint-plugin/test/prefer-separate-type-import.test.mts linguist-generated=true
578589
.config/oxlint-plugin/test/prefer-spawn-over-execsync.test.mts linguist-generated=true

CLAUDE.md

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -47,6 +47,10 @@ Full ruleset + threat model + bypass surface in [`docs/claude.md/fleet/public-su
4747

4848
Full ruleset — open-PR edits, Bugbot inline replies, rebase-over-revert for unpushed commits, no-empty-commits, commit-author canonical identity, scan-label scrubbing, enterprise-ruleset bypass — in [`docs/claude.md/fleet/commit-cadence-format.md`](docs/claude.md/fleet/commit-cadence-format.md).
4949

50+
### Prose authoring (commit bodies, PRs, CHANGELOG, docs)
51+
52+
🚨 Run human-facing prose through the `prose` skill before it lands: commit message bodies, PR descriptions, CHANGELOG entries, README sections, `docs/` markdown. The skill catches throat-clearing openers, "not X, it's Y" contrasts, em-dash chains, adverbs doing vague work, metronomic rhythms. Subject lines stay terse and imperative under `commit-message-format-guard`. Cascade commits and bot output are exempt. Full rules: [`.claude/skills/prose/SKILL.md`](.claude/skills/prose/SKILL.md).
53+
5054
### Squash-history opt-in
5155

5256
Some fleet repos squash the default branch on a cadence — currently socket-addon, socket-bin, socket-btm, sdxgen, stuie (declared via `optIns: ['squash-history']` in `template/.claude/skills/cascading-fleet/lib/fleet-repos.json`). When working in an opted-in repo, prefer one consolidated commit per logical change over a long fan of tiny WIP commits; the `squashing-history` skill is the documented way to collapse history when it grows long. Threshold reminder + bypass `Allow squash-history-reminder bypass` (enforced by `.claude/hooks/squash-history-reminder/`).
@@ -212,6 +216,7 @@ When writing or extending a Bash-allowlist hook, prefer **AST-based parsing** ov
212216
- Before calling done: perfectionist vs. pragmatist views. Default perfectionist absent a signal.
213217
- If a fix fails twice: stop, re-read top-down, state where the mental model was wrong, try something fundamentally different.
214218
- **When the user authorizes a queue** ("complete each one", "hammer it out", "100%", "do them all"): finish every item before stopping. Don't post "what's next?" / "honest stopping point" / "session totals" after one item — that re-litigates intent already given. Continue until the queue is empty or a genuine blocker hits (enforced by `.claude/hooks/dont-stop-mid-queue-reminder/`). Skip AskUserQuestion when recent transcript carries explicit go-ahead directives ("do it" / "yes" / "proceed") — pick the obvious default and execute (enforced by `.claude/hooks/ask-suppression-reminder/`).
219+
- **Direct imperatives → execute, don't litigate.** When the user issues a bare command ("use nvm 26.2.0", "cancel the build", "do it", "kill it"), the response is the tool call, not a paragraph weighing trade-offs. Hedge openers ("That won't help because…", "Let me explain why…", "Before I do that…") + analysis-before-action when the command was unambiguous are the failure mode. State the intent in one short sentence at most, then run the command (enforced by `.claude/hooks/follow-direct-imperative-reminder/`).
215220

216221
### Error messages
217222

pnpm-lock.yaml

Lines changed: 60 additions & 0 deletions
Some generated files are not rendered by default. Learn more about customizing how changed files appear on GitHub.

0 commit comments

Comments
 (0)