decision-mapping

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface within its state-management workflow.
  • Ingestion points: The 'Resume' invocation path loads a project-tracked markdown file (the 'decision map') directly into the session context (SKILL.md).
  • Boundary markers: Absent; the instructions do not provide delimiters or warnings to the agent to treat the markdown content as untrusted data.
  • Capability inventory: The skill can invoke the /prototype skill (which typically involves writing and executing code) and the AskQuestion tool (SKILL.md).
  • Sanitization: Absent; the markdown file is loaded and processed without validation or escaping of its contents.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 08:26 AM
Security Audit — agent-trust-hub — decision-mapping