task-decomposition

Pass

Audited by Gen Agent Trust Hub on Apr 2, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection (Category 8) because it ingests and processes untrusted task descriptions which are then used to guide agent behavior and subagent dispatch.
  • Ingestion points: User-provided project deliverables, task descriptions, and plans used during the decomposition phases (SKILL.md).
  • Boundary markers: Absent; there are no instructions to use delimiters or explicit warnings to the agent to ignore instructions embedded within the processed task data.
  • Capability inventory: The skill explicitly outlines 'Subagent Dispatch Opportunities' using the Agent tool for concurrent execution and integrates with downstream skills like executing-plans and task-management (SKILL.md).
  • Sanitization: Absent; the skill does not define methods for validating, filtering, or escaping content from task descriptions before it is passed to other tools or subagents.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 2, 2026, 12:42 AM