local-work-tracker
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes Feature metadata such as titles and summaries and interpolates them into markdown documentation files like index.md and completed.md. The lack of sanitization for markdown formatting or the skill's own regional markers in these inputs represents an indirect prompt injection surface.\n
- Ingestion points: Feature record files located in .waypoint/tracker/features/ read by the scripts/waypoint_tracker.py utility.\n
- Boundary markers: The script uses explicit and markers to identify updateable regions in markdown files.\n
- Capability inventory: The skill has the capability to read and write files and create directories within the repository root via the Python script.\n
- Sanitization: No sanitization or escaping is performed on Feature titles or summaries before they are written into markdown files.\n- [SAFE]: The skill implements protective measures by automatically updating the repository's .gitignore file to ensure that local identity files and lock files are not committed to version control, reducing the risk of accidental information disclosure.
Audit Metadata