cawplan-my-work

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the CawPlan API via the cawplan CLI tool, creating a surface for potential indirect prompt injection attacks where malicious content in ticket data could influence the agent's behavior.
  • Ingestion points: Ticket titles, types, priorities, status names, history transition fields, progress comments, and linked PR/commit titles/URLs (retrieved in SKILL.md through cawplan todos, cawplan tickets search, and cawplan tickets history commands).
  • Boundary markers: The skill instructions do not specify the use of delimiters or provide explicit warnings to the agent to ignore instructions embedded within the fetched ticket content.
  • Capability inventory: The skill utilizes the Bash tool to execute CLI commands (cawplan) and processes the resulting output to generate reports and summaries.
  • Sanitization: No sanitization, validation, or escaping routines are mentioned for the content retrieved from the external API before it is processed or presented to the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 07:18 AM
Security Audit — agent-trust-hub — cawplan-my-work