verify-feature
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from repository documentation, specifications, and tasks, which represents an ingestion surface for untrusted content.\n
- Ingestion points: The agent is instructed to read repository guidance, requirements, linked tasks, design decisions, and source/test evidence from the working directory.\n
- Boundary markers: There are no explicit delimiters or boundary markers specified for isolating these external inputs from the agent's core instructions.\n
- Capability inventory: The skill allows for running repository-supported commands and bounded local checks, but explicitly forbids modifying the source code, task state, or installing dependencies.\n
- Sanitization: No specific sanitization or filtering of external content is defined, though the skill mandates that source content be treated as evidence rather than authority.\n- [COMMAND_EXECUTION]: The skill instructs the agent to run repository-supported commands and local checks to verify features.\n
- Evidence: The 'Verification execution' section explicitly allows for running bounded local checks while mandating prior inspection of these commands to ensure they do not modify the environment, access live services, or expose secrets.
Audit Metadata