docs: sync README to current contract + surface bench + P11 cross-ref (0.23.2)#70
Conversation
… (0.23.2) Post-ship audit of BRO-1205 (bench MVP) + BRO-1211 (Databricks live mode) found the canonical surfaces current (SKILL.md, references/provider-standards.md, CHANGELOG, spec, KG research entities) but README.md frozen at the P11 era — a CLAUDE.md Self-Documenting Standards rule #3 violation (counts must match SKILL.md, the authoritative source). bstack doctor does not lint README, so this rot was not CI-enforced. README.md: - "Eleven irreducible primitives" → twenty; table extended P1-P11 → full P1-P20 (wording from SKILL.md's enforcement table) - "28 curated skills" → 30 (matches SKILL.md) across intro + Stack-layers header + bootstrap description - Commands: was six (bootstrap/doctor/repair/status/validate/revamp); now also documents bench, wave, crystallize, metrics, skills under an "Orchestration & observability" subsection; bench links to references/provider-standards.md - Reasoning-enforced set corrected (P6, P9-P20) vs mechanism-enforced (P1,P2,P4,P5,P7,P8); closing narrative "eleven" → "twenty" references/primitives.md: - P11 Empirical Feedback Loop now lists `bstack bench` as the dedicated P11 measurement substrate (table row + paragraph), cross-referencing provider-standards.md + the bench spec No code, no behavior change. Companion KG artifact (workspace repo): research/entities/pattern/openai-compatible-provider-abstraction.md. Out of scope (separate audit): deeper skill-count reconciliation (SKILL.md 30 curated vs companion-skills.yaml 65 full roster). README follows SKILL.md per rule #1. Ticket: BRO-1376 Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
Caution Review failedThe pull request is closed. ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (4)
📝 WalkthroughWalkthroughThis PR bumps the release version to 0.23.2 and synchronizes documentation across README, CHANGELOG, and primitives reference to reflect expanded primitive scope (eleven to twenty primitives) and extended command surface with new orchestration capabilities, while explicitly documenting that no code or behavior changed. ChangesVersion 0.23.2 Documentation Synchronization
Estimated code review effort🎯 1 (Trivial) | ⏱️ ~5 minutes Possibly related issues
Possibly related PRs
Poem
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Summary
Post-ship documentation audit (after BRO-1205 bench MVP + BRO-1211 Databricks live mode). The canonical surfaces were already current — SKILL.md, references/provider-standards.md, CHANGELOG, specs/bench-skill-evolution.md, and the KG research entities all document the bench/provider work. The gap was the public README.md, frozen at the P11 era.
This is a docs-only PR. No code, no behavior change.
What changed
README.md— synced to SKILL.md (the authoritative source per CLAUDE.md Self-Documenting Standards rule #3):bench,wave,crystallize,metrics,skillsunder an "Orchestration & observability" subsection (benchlinks toreferences/provider-standards.md)references/primitives.md— P11 Empirical Feedback Loop section now listsbstack benchas the dedicated P11 measurement substrate (validation-surface table row + paragraph), cross-referencingprovider-standards.md+ the bench spec.CHANGELOG.md+VERSION— 0.23.1 → 0.23.2 docs patch release (matches the established docs-patch convention; 0.23.1 was itself adocs:release).Why
bstack doctordoes not lint the README, so this drift was invisible to CI — pure quality debt. The "eleven primitives" headline in a 20-primitive workspace is a rule #3 violation, andbench(BRO-1205/1211) was one of 5 commands missing from the public front door.Companion artifact (workspace repo, not this PR)
research/entities/pattern/openai-compatible-provider-abstraction.md— P6 KG entity documenting the reusable provider-abstraction architecture BRO-1211 introduced (lint-clean).Out of scope
Deeper skill-count reconciliation: SKILL.md says 30 curated;
companion-skills.yamllists 65 (full roster incl. optional). README follows SKILL.md per rule #1; the 30-vs-65 question is a separate audit.Validation
grep -niE "eleven|28 skills|six commands" README.md→ cleanbash scripts/doctor.sh --quiet→ 96/97 (the 1 gap is a pre-existingcompute-budget-statusitem, unrelated)Ticket: BRO-1376
🤖 Generated with Claude Code
Summary by CodeRabbit
Documentation
wave,crystallize,metrics,skills,bench