200-agents-md
Pass
Audited by Gen Agent Trust Hub on Jul 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a strictly defined, multi-step interactive workflow that requires explicit user input and confirmation for each piece of data documented in the output file.
- [COMMAND_EXECUTION]: The skill identifies and documents standard development commands such as Maven wrapper operations (e.g.,
./mvnw). It does not execute these commands autonomously, but rather records them for documentation purposes. - [PROMPT_INJECTION]: The skill analyzes existing repository context and documentation to inform its discovery phase. This ingestion is moderated by an interactive session that requires the user to validate or provide specific answers, providing a natural defense against indirect injection from project files.
- Ingestion points: SKILL.md (Workflow Step 1: Inspect project structure and existing documentation).
- Boundary markers: The mandatory step-by-step question format acts as a structural boundary between ingested data and final output.
- Capability inventory: File system access for reading project structure/documentation and writing the AGENTS.md file.
- Sanitization: Content is filtered through a user-controlled question-and-answer process before being written to the repository.
Audit Metadata