tree-of-thoughts

Pass

Audited by Gen Agent Trust Hub on Apr 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill performs basic command execution to create local directories (.specs/research and .specs/reports) for managing its internal reasoning artifacts. This is a standard and safe operation restricted to the project environment.
  • [SAFE]: The skill processes external task descriptions and context by passing them to subagents. It mitigates the risk of indirect prompt injection by using explicit XML-style boundary markers (e.g., <task>, <context>) in its prompt templates to separate instructions from user-provided data.
  • [SAFE]: The multi-agent orchestration and adaptive strategy selection (SELECT_AND_POLISH, REDESIGN, FULL_SYNTHESIS) are implemented as logic for the agent to follow, involving no dynamic code execution (eval) or remote script fetching.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 23, 2026, 03:49 AM