Evidence · DiggingBeagle record
PyCache Trap: hidden Python bytecode defeats source-only AI skill scanners
Primary or indexed evidence for Radar finding DB-RADAR-CP2-CP3-012. Controlled scanner tests, no confirmed skill-marketplace infection.
- Source role
- primary disclosure
Evidence record
Primary or indexed evidence for Radar finding DB-RADAR-CP2-CP3-012. Controlled scanner tests, no confirmed skill-marketplace infection.
Claim-level citations (3)
- supportsPyCache Trap exposes a blind spot in agent skill scanners: The PyCache Trap researchers report 94% to 100% bypass success across 100 agent skills and seven scanners in their controlled experiments.
Abstract, scanner-guided rewriting and evaluated skills
- supportsPyCache Trap exposes a blind spot in agent skill scanners: Their execution-aware validation detected 100 of 100 source-present cache substitutions and reported 92.8% recall with 10% false-positive rate in expanded evaluation.
Abstract, execution-aware validation results
- supportsPyCache Trap exposes a blind spot in agent skill scanners: The EAV follow-up study reports 92.8% recall at a 10% false-positive rate across five attack families and 200 benign skills; it does not establish universal loader coverage.
Abstract, expanded EAV evaluation