mapping-mitre-attack-techniques

Pass

Audited by Gen Agent Trust Hub on Jul 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches the authoritative MITRE ATT&CK Enterprise STIX data bundle from the official MITRE GitHub repository.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests and processes external datasets to generate reports and heatmap layers. * Ingestion points: The scripts/agent.py script reads data from enterprise-attack.json and a user-provided detection_rules.json. * Boundary markers: There are no explicit delimiters or 'ignore instructions' warnings implemented in the data ingestion logic. * Capability inventory: The script performs file system read and write operations to load datasets and export Navigator layers. * Sanitization: The skill relies on standard JSON parsing and the official mitreattack-python library for structured data handling.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 16, 2026, 09:12 AM
Security Audit — agent-trust-hub — mapping-mitre-attack-techniques