review-work

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it ingests untrusted data from task descriptions and acceptance criteria and has the capability to perform write operations in the project management system.\n
  • Ingestion points: Untrusted external data is loaded via the get_work_unit and get_task MCP tools in SKILL.md.\n
  • Boundary markers: Absent; the skill lacks specific delimiters or instructions to treat task-derived data as non-executable text.\n
  • Capability inventory: The skill utilizes write-access tools including update_task, update_work_unit, batch_create_tasks, and create_task_comment to modify project states in AgiFlow.\n
  • Sanitization: Absent; no sanitization, escaping, or validation logic is applied to the ingested content before the agent evaluates it against the audit criteria.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 12:10 AM
Security Audit — agent-trust-hub — review-work