math-olympiad
Solve competition math (IMO, Putnam, USAMO) with adversarial verification and calibrated abstention.
🛡️ AgentReady threat assessment
MAESTRO 7-layer threat model + OWASP AIVSS risk score for math-olympiad, derived from its capabilities.
AIVSS 5.2 · Medium
View MAESTRO 7-layer threat model →Overview
Anthropic's official math-olympiad plugin. It solves competition-level math with fresh-context adversarial verifiers that attack proofs using specific failure patterns, catching errors self-verification misses, and prefers calibrated abstention over bluffing. Surface is subagents/skills that generate and adversarially check proofs.
Key features
- Adversarial fresh-context proof verifiers
- Targets specific proof failure patterns
- Calibrated abstention over bluffing
- Covers IMO, Putnam, USAMO
Use cases
- Attempt hard competition math problems
- Get proofs adversarially checked for gaps