skills/maragudk/fabrik/gardeners/Gen Agent Trust Hub

gardeners

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security issues were detected. The skill uses standard agent orchestration patterns to coordinate sub-agents through a shared task list.
  • [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection as sub-agents ingest external data (code or issues) through the referenced garden skill.
  • Ingestion points: Gardeners scan project files or issue trackers to identify tasks.
  • Boundary markers: Absent; the prompt template does not specify the use of delimiters or 'ignore' instructions for the scanned data.
  • Capability inventory: The skill uses TeamCreate, Agent, TaskList, TaskUpdate, and SendMessage for orchestration; the sub-agents likely possess file-write capabilities via the garden skill.
  • Sanitization: No sanitization or validation of the scanned content is mentioned in the orchestration logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 07:56 AM