pluto-review

name	pluto-review
description	Pluto-specific code review guidelines. Use as a general guideline when asked to conduct a code review.

Principles:

• Functional equivalence first; document and justify deviations.
• Evidence-based: prefer tests, outputs, and file/line references over guesses.
• Minimal change bias; avoid scope creep.
• No time estimates in review output.

When producing a review, include:

1. Summary (1–3 sentences)
2. Findings (ordered by severity)
3. Parity matrix (if applicable)
4. Tests (run or not run)
5. Open questions/assumptions

Severity model:

• Critical: breaks contract, security issue, incompatible output/protocol.
• High: user-visible regression or parity gap with operational impact.
• Medium: behavioral difference with limited impact or edge cases.
• Low: minor inconsistency or optional improvement.

Findings format (use path:line references, 1-based):

- [Severity] Title
  Impact: ...
  Evidence: pluto/crates/foo/src/lib.rs:123
  Go reference: charon/cmd/foo.go:456
  Recommendation: ...

Parity matrix template:

Component	Go	Rust	Match	Notes
CLI flag --foo	present	present	yes
Error string for missing key	"..."	"..."	no	mismatch in punctuation
Wire format	pbio	pbio	yes