living-plan
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines an 'Intake Loop' where data from external reports is ingested into project documentation. This creates a potential surface for indirect prompt injection if malicious instructions are embedded in reports.
- Ingestion points: Processes output from external tools like repo-deep-dive and audit as described in SKILL.md.
- Boundary markers: The convention lacks explicit delimiters for external content, relying on the plan-intake skill's manual approval process.
- Capability inventory: Uses Read, Write, Edit, and Bash to manage and modify project files.
- Sanitization: Incorporates a 'human approval' step to validate proposed entries before they are committed.
Audit Metadata