project-plan
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill follows best practices for project management by researching existing context before proposing changes. It utilizes specific tools (get_project_detail, list_tasks, list_members) to perform its functions.
- [DATA_EXPOSURE]: The skill reads codebase patterns and project metadata to inform planning. This is standard behavior for a development-focused AI agent and does not involve exfiltration to external domains.
- [COMMAND_EXECUTION]: All actions are performed through defined MCP tools for task management and project research. No arbitrary shell command execution or script spawning was detected.
- [PROMPT_INJECTION]: While the skill processes user-provided requirements (an indirect prompt injection surface), it includes explicit guardrails and a clarification step (Step 2) to identify ambiguous details and ask follow-up questions before taking action.
Audit Metadata