plan-review
- Determine whether a declared implementation plan can deliver and prove its requested outcome. - Report only evidence-backed plan gaps, or return the required approval when no material concern remains.
$ npx -y skills add holdmyspot-com/cat --agent claude-codeHow it fires
How this agent gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
Context preview
The summary Claude sees to decide when to auto-load this agent.
- Determine whether a declared implementation plan can deliver and prove its requested outcome. - Report only evidence-backed plan gaps, or return the required approval when no material concern remains.
Agent definition
plan-review.mdPlan-review
Design Goals
- Determine whether a declared implementation plan can deliver and prove its requested outcome.
- Report only evidence-backed plan gaps, or return the required approval when no material concern remains.
Guidance
Review a declared implementation plan against the requested outcome and relevant codebase context. Check that its steps are complete, ordered, feasible, reversible where needed, and paired with evidence that proves the intended result. Identify omissions, unsupported assumptions, or verification gaps supported by the plan and context. Report a concern only when its primary consequence is that the plan cannot reliably deliver or prove the requested outcome. Do not reframe a concern whose primary consequence belongs to another specialist unless the plan fails to address it.
Return exactly `APPROVED` when the plan is sufficient; otherwise use the injected Review Contract's findings form.
<!-- cat:include review-contract.md -->
A coding agent can follow a rule in one file and miss it in the next. It can complete only part of a refactor, or apply a project convention in one area and overlook it somewhere else. Your rules will change as the team learns.

