adversarial testing agents

smshahbaj/crucible

Agent

Performs a cheap first-pass triage to decide whether a decision needs no extra review, a small independent review, or deep review. Use before spawning other specialists.

smshahbaj/crucible

Agent

Produces the final evidence-weighted decision after independent review, targeted challenge, and optional red-team. It does not majority-vote and it does not reveal hidden chain-of-thought.

smshahbaj/crucible

Agent

Performs a compact pre-final quality gate for consequential decision reviews. It detects unsupported leaps, anchoring, consensus illusion, scope mismatch, reversibility blindness, base-rate neglect, single-point dependencies, false precision, and unnecessary continuation.

alejandrosaenz117/bonfires-marketplace

Agent

Use this agent to challenge code reviews, architecture decisions, or security approvals by finding the most plausible failure mode. The Devil's Advocate identifies where systems collapse.

tasumermaf/the-adversary

Agent Claude Code

FIND lens (Round 2) — refutes the claim that every promise in the abstract/summary is delivered, in matching strength, by the body.

tasumermaf/the-adversary

Agent Claude Code

FIND lens (Round 5) — refutes the claim that the prose is free of AI-generated verbal tics that signal unedited machine writing.

tasumermaf/the-adversary

Agent Claude Code

FIND lens (Round 1 / release gate) — refutes every number in the artifact by RECOMPUTING it from canonical data. Reading a summary is never verification.