kata-insert-phase
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill takes user input and writes it into documentation files, which could be used to influence the agent's future behavior or disrupt project state.\n
- Ingestion points: User-provided phase description arguments in
SKILL.md.\n - Boundary markers: Absent. The description is interpolated directly into the markdown content of
ROADMAP.mdandSTATE.md.\n - Capability inventory: The skill uses bash commands to create directories and update documentation files, and invokes a local Node.js utility script.\n
- Sanitization: Incomplete. The directory slug is sanitized using
sed, but the description text is not sanitized before being written into markdown files.
Audit Metadata