implement
Pass
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes external specifications (PRDs, issues) to generate tests and code, creating a surface for indirect prompt injection.
- Ingestion points: External PRDs, specification files, or linked issues referenced in Step 1.
- Boundary markers: Absent; external content is processed without explicit delimiters or instructions to ignore embedded prompts.
- Capability inventory: Shell execution of testing frameworks (npm, pytest, cargo) and git version control operations.
- Sanitization: Absent; content from specifications is used directly to define implementation tasks.
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to run test suites, type checkers, and git operations.
- Evidence: Commands such as
npm test,pytest,npx tsc, andgit commitare used throughout the 5-step loop.
Audit Metadata