skill-plan

Pass

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted external data (requirements documents) to generate implementation plans, which constitutes a surface for indirect prompt injection.
  • Ingestion points: Requirements documents (REQ-*.md) and chat history summaries in SKILL.md.
  • Boundary markers: The instructions do not define delimiters or specific "ignore" commands to separate user-provided requirements from the agent's core instructions.
  • Capability inventory: The skill is limited to read-only tools including codebase exploration (Glob), pattern searching (Grep), and file reading (Read). No file modification or network tools are invoked.
  • Sanitization: There is no evidence of input validation, escaping, or instruction filtering for the ingested requirements.
  • [NO_CODE]: The skill contains no executable scripts or code files, consisting entirely of markdown instructions and configuration.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 17, 2026, 03:28 AM
Security Audit — agent-trust-hub — skill-plan