unity-implement-batch

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-supplied task definitions and specifications from local project directories, which constitutes an attack surface for instructions embedded in documentation.
  • Ingestion points: The skill reads files from docs/<feature-slug>/tasks and their referenced specifications as the primary source for subagent instructions.
  • Boundary markers: Section 1 defines specific gates (presence of CLI tools and UI modules) that must be satisfied before delegation, and Section 2 requires subagents to follow established project standards.
  • Capability inventory: The skill utilizes /unity-cli for modifying Unity-specific files (scenes, prefabs, ScriptableObjects) and /unity-tdd for executing automated behavioral tests.
  • Sanitization: The workflow includes a remediation stage in Section 4 that explicitly requires the agent to verify review findings against the current repository state and classify them as valid or false positives rather than applying them blindly.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 11:53 PM