plan-project
Pass
Audited by Gen Agent Trust Hub on Apr 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: Uses the Bash tool to create the 'docs' directory using the command 'mkdir -p docs'.
- [PROMPT_INJECTION]: The skill processes untrusted project requirement documents (PRD) which creates a surface for indirect prompt injection. 1. Ingestion points: Searches for and reads 'prd.md' files located in the project structure. 2. Boundary markers: Absent; file paths are provided to sub-agents without instructions to ignore embedded commands. 3. Capability inventory: The skill and its specialists have access to the 'Write' tool for creating markdown files and the 'Task' tool for spawning further sub-agents. 4. Sanitization: No content validation or escaping is performed on the input PRD files before they are read by sub-agents.
Audit Metadata