using-superpowers

Pass

Audited by Gen Agent Trust Hub on Oct 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill acts as a coordinator for other workflow skills and provides platform-specific tool mappings for Codex, Gemini CLI, Hermes Agent, Antigravity, and Pi.
  • [SAFE]: Includes explicit safety guidelines in SKILL.md instructing the agent to treat untrusted repository examples and external content as data, not instructions, which is a best practice for preventing indirect prompt injection.
  • [SAFE]: Tool mappings for subagent dispatch (e.g., spawn_agent, invoke_agent, delegate_task) use standard, platform-provided mechanisms for multi-agent coordination.
  • [SAFE]: Environment detection logic provided in the Codex reference file uses standard read-only git commands to identify the workspace state without making unauthorized changes.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 18, 2026, 12:45 PM