executing-plans
Pass
Audited by Gen Agent Trust Hub on Jun 21, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized network activity detected.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon instructions from external plan files (e.g., docs/plan.md). This represents a potential surface for indirect prompt injection if the plan file contains malicious instructions.
- Ingestion points: Plan files are read in Step 1, Task 1.
- Boundary markers: No technical delimiters are used to separate plan content, but the skill mandates a critical review process.
- Capability inventory: The skill can perform git operations (commit, add), run test commands (npm test, pytest), and write to task lists (TodoWrite).
- Sanitization: The skill explicitly instructs the agent to identify missing dependencies or environment assumptions and to pause for human approval if doubts arise.
Audit Metadata