code-review-burntsushi

name: code-review-burntsushi description: Review code through BurntSushi's "honest machinery" lens — asking whether APIs, invariants, error contracts, performance boundaries, and defaults accurately represent reality. Use when reviewing library APIs, patches that change complexity or cost, dependency additions, or when the user asks "is this API honest?", "will this panic?", "is the cost visible?", "does this scale?", or "are the defaults right?". Works across languages; see REFERENCE.md for Rust-specific elaborations. Do NOT use for deletion/anti-bloat review (use code-review-russ-cox) or product-vs-library tradeoffs (use code-review-mitsuhiko).

"Is there a reasonable expectation that the code will behave as advertised?" — BurntSushi

A code review should ask: is this code honest about its invariants, failure modes, costs, and semantics? Not just: does it work?

Quick start

Paste a diff or file and say "burntsushi review" or "honest review". Walk the ten general axes in REFERENCE.md, add the two Rust-specific axes if the code is Rust, then emit findings.

Steps

State the change — what is this patch doing in one sentence? If unclear, ask first.
Walk all ten general axes from REFERENCE.md in order. Flag every axis with a finding.
If Rust: also apply the two Rust-specific axes (encoding invariants, ownership/allocation).
Emit findings first using the output contract below.

Output Contract

Lead with findings, ordered by severity. Use Must fix, then Consider, then Open questions; omit empty sections.
Prefix findings with [Axis N | file:line].
Each finding names the invariant, cost, or failure mode and gives one concrete alternative.
Add a brief Summary after findings. Do not add praise unless the user asks.

Tone rules

Name the invariant, failure mode, or hidden cost precisely.
Suggest one practical alternative per finding.
Do not nitpick style, naming, or formatting unless it obscures correctness or cost.
Do not complain about unwrap/assertions on static programmer invariants — only flag panics when the failure source is runtime/user/environment.

Reference

REFERENCE.md — all axes with examples, model comments, and skill rules.