pm-roadmap-add
Pass
Audited by Gen Agent Trust Hub on Jun 20, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill implements a mandatory 'Input Threat Scan' (Step 1.1) to detect malicious patterns in user-supplied arguments. Furthermore, the 'Safety Rules' explicitly instruct the agent to treat all external content from PRDs, stories, and the codebase as data only, ignoring any instruction-like text found within those files.
- [DATA_EXFILTRATION]: While the skill scans the project environment for context, it enforces a strict exclusion list in its 'Safety Rules' and 'Step 2' for sensitive files, including .env, secrets, .ssh, .pem, and .key files. Data extraction is restricted to file names and summary counts (e.g., TODO/FIXME counts) rather than full file contents.
- [COMMAND_EXECUTION]: The skill uses restricted shell commands for standard version control operations (git add, git commit) and for retrieving the repository URL (git remote get-url) to generate issue links. These operations are limited in scope and integrated into a structured workflow that requires human confirmation.
- [SAFE]: A 'HARD STOP' is implemented after Phase 1, ensuring that the roadmap item and its metadata are previewed and approved by the user before any file modifications or git commits are executed in Phase 2.
Audit Metadata