Cavemenko review
One-line code review comments. Use when user asks to review code, check changes, or calls /cavemenko-review.From its SKILL.md
npx -y skills add ruslanlap/cavemenko --skill cavemenko-reviewAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 4 stars4 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
1.3 KB, as published. Nobody here has run it
Review current code changes. Format — one-line comments.
Format
One line per issue:
L<line_number>: <severity> <problem>. <fix>.
Severity:
bug— broken, doesn't work as intendedrisk— works but dangerous (race, leak, security)nit— style, minorq— question, needs clarification
Rules
- Skip praise — "nicely written", "great idea" not needed
- Skip obvious —
missing semicolon,trailing whitespace— linter's job - If code ok — say
LGTMand stop - Max one line per issue; if more needed — split into multiple
- Line reference required
Examples
Good:
L42: bug off-by-one in loop, `i <= arr.length` → IndexError. Use `<`.
L78: risk `JSON.parse` without try/catch — crash on invalid input.
L104: nit magic constant `86400`, replace with `SECONDS_PER_DAY`.
L130: q why retry with delay=0? This is a busy loop.
If all clean:
LGTM
Bad:
L42: Nice function, but could be improved. Maybe worth thinking about...
Process
- Read diff (
git diffor specified file) - For each issue — one line in format above
- Sort by severity: bug → risk → nit → q
- If no issues —
LGTMand stop