polecat-worker

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is designed to autonomously execute instructions contained within 'beads' (work items). \n
  • Ingestion points: The agent retrieves work items via state.getHook(myAgentId) in SKILL.md, specifically accessing workItem.beadId, workItem.title, and the work item description which provides the execution context. \n
  • Boundary markers: There are no explicit boundary markers or instructions provided to the agent to disregard potentially malicious commands embedded within the task descriptions. \n
  • Capability inventory: The skill possesses capabilities for file system access via a StateManager (reading and writing in .chipset/state/) and execution of git commands (git checkout, git commit, git push) as seen in SKILL.md and references/examples.md. \n
  • Sanitization: No evidence of input validation or sanitization is present for the data retrieved from the work items before it is used in logic or command execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 08:34 PM
Security Audit — agent-trust-hub — polecat-worker