6 frameworks. Every rule mapped.
Every ATR rule carries mapping metadata for OWASP, MITRE ATLAS, NIST AI RMF, EU AI Act, ISO 42001, and SAFE-MCP. All metadata is MIT-licensed, downloadable, and auditable.
Compliance officers cannot submit URLs as evidence in procurement. Download the structured compliance mapping package (PDF + JSON) with per-rule framework mappings, rule ID index, and quality score summary.
Full coverage across all 10 agentic AI risk categories.
All 113 ATR rules carry MITRE ATLAS technique references. Grouped by tactic in the rule explorer.
All rules carry NIST AI RMF subcategory mappings. 16 subcategories across GV/MP/MS/MG. OSCAL catalog accepted under Path 1.
78 of 85 techniques covered (OpenSSF MCP security framework).
Rules map to high-risk system obligations (Art. 9, 10, 15) for AI systems deployed in agentic contexts. Mapping documented per rule.
Rules map to AI management system controls for risk identification, monitoring, and incident response.
ATR's compliance mappings are not marketing claims. Each rule's YAML contains specific compliance metadata citing the exact regex or token that detects the attack — not a generic claim of 'alignment with framework'.
Browse raw rule YAML on GitHub →