project-workflow-analysis-blueprint-generator
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill generates instructions for an agent to analyze a project's codebase, which constitutes untrusted input data.
- Ingestion points: The prompt template in
SKILL.mddirects the agent to read and document various parts of a codebase, including API controllers, service layers, and data access implementations. - Boundary markers: The generated prompt does not include delimiters or specific instructions to the agent to ignore or isolate any natural language instructions that might be embedded within the source code comments or strings of the project being analyzed.
- Capability inventory: To perform the requested task, the agent typically requires read access to the entire project filesystem.
- Sanitization: No methods for sanitizing, filtering, or validating the content of the ingested source code are included in the skill's instructions.
Audit Metadata