Expand CONTRIBUTING guidelines and audit them in the review skill#4493
Merged
Conversation
twmb
force-pushed
the
contributing-and-review-skill
branch
2 times, most recently
from
2f61530 to
0f3c0d4
Compare
…nd PR size limit - §1.3.4: split benchmarking into two required phases (local/localhost and real-endpoint), pointing at docs/benchmarking.md for the standard process and docs/benchmark-results/ for recording results. - §3.1.2/§3.1.3: require gofumpt formatting and add a Go-first rule — connectors are idiomatic Go built to support their endpoint well (per the rest of this document), never a line-by-line port of another ecosystem's implementation (e.g. Debezium for CDC). - §3.3: add a contribution-process section keeping commits/PRs reviewable, with a ~10K-line guideline on code a reviewer must read (AI-generated code included); excludes generated/derived files, vendored code, and non-code (docs, skills, Terraform); large features split into smaller PRs. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Add a Certification & Contribution Guidelines agent that walks all of CONTRIBUTING.md and delegates section ownership to the existing specialist agents (Go-first -> godev, tests -> tester, benchmarking -> benchmarking agent, change size -> commit policy) to avoid duplicate findings. Load CONTRIBUTING.md into review context, teach godev to flag §3.1.3 foreign ports and commit policy to flag the §3.3 size guideline (counting code a reviewer must read, including AI-generated, and skipping generated/non-code-heavy PRs), and widen the filter to allow CONTRIBUTING.md findings across all sections (subjective criteria only on a clear, material gap; §2 selection criteria as non-blocking notes). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
twmb
force-pushed
the
contributing-and-review-skill
branch
from
0f3c0d4 to
edda777
Compare
squiidz
approved these changes
Jun 12, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Extends the connector certification & contribution guidelines and wires the
/reviewskill to enforce them.CONTRIBUTING.mddocs/benchmarking.mdfor the standard process anddocs/benchmark-results/for recording results.gofumptformatting, and add a rule that connectors are idiomatic Go built toward a goal — supporting the target endpoint well, where "well" is defined by the rest of the document (UX, configuration knobs/validation, observability, reliability). A reference implementation (e.g. Debezium for CDC) may be consulted to understand the endpoint, but it is not a spec to replicate and we are not bound to its abstractions, idioms, or naming — never a line-by-line port..claude/skills/review/SKILL.mdCONTRIBUTING.mdinto review context (applies to internal and external contributions).godev, tests →tester, benchmarking → benchmarking agent, change size → commit policy) so there's no duplicate flagging.godevto flag §3.1.3 foreign ports and the commit-policy agent to flag the §3.3 size guideline — counting code a reviewer must read (incl. AI-generated), and skipping generated/non-code-heavy PRs.Why
Codifies internal contribution standards and makes review enforce them automatically for both internal and external contributors.
Scope note
Benchmarking infrastructure harmonization (reconciling
docs/benchmarking.mdwith the in-flight AWS bench framework in #4487, and teaching the review skill aboutbenchmarking/aws/) is intentionally out of scope here — that's left to the benchmarking work in #4487.🤖 Generated with Claude Code