home / skills / yeachan-heo / oh-my-claudecode / review

review skill

/skills/review

This skill evaluates and provides actionable feedback on existing plans by triggering Critic review, speeding up approvals.

npx playbooks add skill yeachan-heo/oh-my-claudecode --skill review

Review the files below or copy the command above to add this skill to your agents.

Files (1)
SKILL.md
683 B
---
name: review
description: Alias for /plan --review
---

# Review (Plan Review Alias)

Review is a shorthand alias for `/oh-my-claudecode:plan --review`. It triggers Critic evaluation of an existing plan.

## Usage

```
/oh-my-claudecode:review
/oh-my-claudecode:review "path/to/plan.md"
```

## Behavior

This skill invokes the Plan skill in review mode:

```
/oh-my-claudecode:plan --review <arguments>
```

The review workflow:
1. Read plan file from `.omc/plans/` (or specified path)
2. Evaluate via Critic agent
3. Return verdict: APPROVED, REVISE (with specific feedback), or REJECT (replanning required)

Follow the Plan skill's full documentation for review mode details.

Overview

This skill is a shorthand alias that runs a plan review workflow by invoking the Plan skill in review mode. It provides a quick way to trigger a Critic evaluation of an existing plan and returns a clear verdict and actionable feedback. Use it to validate project plans stored in the expected plans directory or a custom path.

How this skill works

When invoked, the skill calls the Plan skill with the --review flag and the provided arguments. It reads the target plan file (defaulting to .omc/plans/ or an explicit path), runs a Critic agent evaluation, and produces a verdict: APPROVED, REVISE (with concrete feedback), or REJECT (replanning needed). The output focuses on correctness, gaps, risks, and recommended changes.

When to use it

  • Before executing a multi-agent plan to catch design or coordination issues early.
  • When you need an independent Critic assessment of completeness and risk.
  • During code review cycles to validate plan assumptions and task breakdowns.
  • When iterating on a plan and you want targeted revision guidance.
  • To gate deployment or execution until plans meet defined quality criteria.

Best practices

  • Store plans under .omc/plans/ so the skill can find them by default.
  • Provide an explicit path when reviewing a nonstandard plan location.
  • Address 'REVISE' feedback with concrete edits and re-run the review.
  • Use the verdict to decide whether to proceed, iterate, or replan.
  • Combine with other tooling for automated checks and CI gating.

Example use cases

  • Run a quick review of a drafted plan before starting parallel agent execution.
  • Integrate the review alias into a CI job that blocks execution on REJECT.
  • Ask the Critic to evaluate a plan after major scope changes to surface regressions.
  • Use the review output to create targeted tickets for required revisions.
  • Validate task sequencing and dependencies in a complex multi-agent orchestration.

FAQ

What verdicts does the review return?

It returns APPROVED, REVISE (with specific feedback), or REJECT (indicating replanning is required).

Where does it look for plans by default?

By default it reads plan files from the .omc/plans/ directory unless you supply a specific path.