convention-over-configuration
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill directs the agent to ingest content from the .agents/stack-context.md file to determine the project stack.
- Ingestion points: SKILL.md in the Before Applying section.
- Boundary markers: There are no instructions to the agent to treat the content of the external file as untrusted or to ignore embedded instructions.
- Capability inventory: The skill references using the context7 MCP tool and web search capabilities.
- Sanitization: No sanitization or validation logic is defined for the external context file.
Audit Metadata