do-in-steps

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns detected. The skill provides a standard procedural framework for task decomposition and sequential agent orchestration.
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the flow of user-provided task descriptions and intermediate outputs through a chain of sub-agents. This orchestration creates an inherent processing surface for external data, but the skill follows best practices such as summarizing context to prevent over-propagation and assigning specific scopes to agents to limit impact.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 01:53 PM
Security Audit — agent-trust-hub — do-in-steps