maintain-verification-skill
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external project data including source code and feature maps to identify documentation drift. This ingestion represents a surface for indirect prompt injection, as the agent processes untrusted project files to generate summaries and recipes. Evidence Chain: (1) Ingestion points: Project source files and feature documentation are read in Step 2. (2) Boundary markers: Absent. (3) Capability inventory: The skill performs file reads/writes and local script execution. (4) Sanitization: Absent. Evidence in SKILL.md (Steps 2 and 4).
- [COMMAND_EXECUTION]: The skill is designed to execute project-local 'harness scripts' to drive application features during its live verification pass. This execution is a primary component of its stated verification purpose. Evidence in SKILL.md (Steps 4 and 5).
Audit Metadata