check
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs a subagent to execute the project's test suite to verify implementation status. This involves running shell commands or scripts within the local development environment.
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from the codebase and plan files, creating a potential vector for instructions embedded in project files to influence agent behavior.
- Ingestion points: Reads plan files in
.specs/plans/and arbitrary modules across the codebase during the review phase. - Boundary markers: No specific delimiters or instructions are provided to the subagent to treat implementation code as data rather than instructions.
- Capability inventory: Includes reading project files, writing to plan files, and executing test suites.
- Sanitization: No sanitization or filtering mechanisms are defined for the data ingested from the codebase.
Audit Metadata