write-plan

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill operates entirely within a restricted scope (.beagle/concepts/) and enforces a strict workflow that requires a source specification file (spec.md) before proceeding.
  • [SAFE]: It utilizes a 'hard gate' to prevent the model from generating plans based on unverified assumptions, routing the user to a brainstorming skill instead.
  • [SAFE]: The skill implements a robust self-review checklist and optional subagent review process to ensure plan quality and adherence to project conventions (CLAUDE.md).
  • [SAFE]: File writing is protected by a manual approval gate; the model must present a draft and receive explicit user confirmation before saving the plan to disk.
  • [SAFE]: The skill strictly separates planning from execution, explicitly stating it does not execute code, run tests, or modify production files beyond the specific plan document.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 04:40 PM
Security Audit — agent-trust-hub — write-plan