/review-pr

Run the canonical llmwiki code review against a pull request and post findings.

Usage: /review-pr <pr-number-or-url>

This slash command loads `docs/maintainers/REVIEW_CHECKLIST.md` and applies every section to the diff of the given PR. The rule is: one concern per bullet, blocker-vs-nit classification, link to the relevant doc section for every finding.

Workflow

1. **Resolve the PR number** — accept either `123` or `https://github.com/Pratiyush/llm-wiki/pull/123`. Use `gh pr view <N>` to fetch the title, body, author, and file list.

2. **Read the governance docs first** — load all of:

• `docs/maintainers/REVIEW_CHECKLIST.md` (the full bar)
• `docs/maintainers/ARCHITECTURE.md` (layer boundaries)
• `docs/maintainers/DECLINED.md` (prior rejections that may apply)
• `CONTRIBUTING.md` (TL;DR rules)
• `SECURITY.md` (privacy + security bar)

3. **Fetch the diff** — `gh pr diff <N>` for the raw patch. Walk the diff file by file, mapping each to its layer via `docs/maintainers/ARCHITECTURE.md`.

4. **Run the checklist** — apply every section of `REVIEW_CHECKLIST.md` to the diff:

• Meta (linked issue, one concern, commit title, CHANGELOG, tests, CI)
• Layer boundaries (layer-appropriate changes, no new runtime deps)
• Security + privacy (no real session data, XSS, no network,

localhost binding, no telemetry)

• Code quality (docstrings, comments, error handling, type hints,

dead code)

• Tests (happy + edge, descriptive names, regression, tmp_path,

run locally)

• Docs (README, CHANGELOG, docs/, docstrings)
• Build + runtime smoke

5. **Classify each finding** — two categories only:

• **Blocker** — Meta section failures, security/privacy issues,

failing tests, failing build, broken layer boundaries

• **Nit** — style, wording, missing comment, cosmetic

6. **Post the review** — group findings by file, then by severity within each file. Blockers first. Every finding cites the relevant section of `REVIEW_CHECKLIST.md` so the contributor can look up the rule.

7. **Final verdict** — one of:

• **Approve** — no blockers, maybe a few nits
• **Request changes** — at least one blocker
• **Comment** — questions only, no blockers, no nits (rare)

Output format

# Review: PR #123 — <title>

**Author:** @<name>
**Verdict:** <Approve | Request changes | Comment>
**Summary:** <one sentence>

## Blockers

### `llmwiki/build.py`

- **[Security]** `<finding>` — see [REVIEW_CHECKLIST §Security + privacy](docs/maintainers/REVIEW_CHECKLIST.md#security--privacy)

### `tests/test_foo.py`

- **[Tests]** `<finding>` — see [REVIEW_CHECKLIST §Tests](docs/maintainers/REVIEW_CHECKLIST.md#tests)

## Nits

### `docs/bar.md`

- **[Docs]** `<finding>` — optional

## Next steps

<!-- What the contributor should do to get to Approve -->

Blocker shortlist

If any of these are true, mark as **Request changes** immediately:

• Tests are failing on master (CI red)
• No linked issue and the PR touches > 2 files
• New runtime dependency introduced in `pyproject.toml`
• Real session data committed under `tests/fixtures/` or `wiki/`
• The maintainer's real username appears anywhere in committed files
• `CHANGELOG.md` is not updated for a `feat:` / `fix:` PR
• Network calls added to `build.py` or `convert.py`
• Server binds to `0.0.0.0` by default

Append to log

After posting the review, append one line to `wiki/log.md`:

[YYYY-MM-DD] review | PR #<N>: <verdict>