simplify-codebase
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze existing source code and tests, which could contain malicious instructions designed to influence the agent's behavior during the refactoring process.
- Ingestion points: Workflow steps 2, 3, and 5 in SKILL.md require the agent to read existing codebase modules and tests.
- Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded prompts within the source code being analyzed.
- Capability inventory: The skill involves codebase analysis, architectural design, and staged code implementation, which assumes the agent has file-system read and write capabilities.
- Sanitization: The workflow implements a form of 'clean room' design in Step 4, where the design phase is performed by an isolated agent given only a behavioral contract and no direct access to the original code, significantly mitigating the risk of instructions leaking from the old code into the new design.
Audit Metadata