The Agent Skills Directory

[SAFE]: The skill is designed for incident analysis and does not contain any malicious code or unsafe operations. Logic is limited to keyword mapping and report generation.
[INDIRECT_PROMPT_INJECTION]: The skill has a potential surface for indirect injection as it processes incident descriptions from external reports.
Ingestion points: scripts/agent.py (processes event descriptions in analyze_incident function).
Boundary markers: None present to distinguish data from instructions.
Capability inventory: None. The skill does not use network, file system, or shell capabilities.
Sanitization: None. Data is interpolated directly into the report.
[EXTERNAL_DOWNLOADS]: References to external services like MITRE ATT&CK Navigator in documentation are for legitimate integration and target a well-known service.

analyzing-cyber-kill-chain