specrail-plan-tasks
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a project-local script
checks/route_gate.pyto verify that the implementation route is ready before creating tasks. - [PROMPT_INJECTION]: The skill processes user-provided specifications from
specs/GH<issue-number>/product.mdandspecs/GH<issue-number>/tech.md, creating an indirect prompt injection surface. 1. Ingestion points: The skill reads markdown files located in thespecs/directory. 2. Boundary markers: The skill body contains explicit instructions to prevent the agent from implementing tasks or removing human approval gates. 3. Capability inventory: The skill has access to read and write files within the project structure and execute a specific Python script. 4. Sanitization: No explicit sanitization or content validation for the specification documents is mentioned.
Audit Metadata