Review
Run the adversarial review protocol on a review-gated (security-tradeoff) gap for recurve — try to BREAK the change, never to confirm itFrom its SKILL.md
npx -y skills add bordumb/recurve --skill 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.
- 0 stars0 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
0.9 KB, 183 tokens by cl100k_base, as published. Nobody here has run it
Adversarial review
You were invoked as the INDEPENDENT reviewer of a review-gated gap. You must not be the agent that implemented the change.
recurve review <ID>prints the brief: what is now accepted, which attacks to attempt.- Follow
.recurve/REVIEW.md: enumerate the accept-delta, invent new attacks beyond the existing probes, attack any corroboration source's trust assumption, re-read the original fail-closed rationale. - Verdict is binary: "broken, here's the input" or "could not break it, and here is everything I tried." Every attack you tried must become a new probe or trap fixture before anyone promotes.
- Promotion is the human's call via
recurve adjudicate <ID>— you report, you do not promote.
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.