Skill

reviewing-code

Runs codex review CLI on git worktree changes, parses results into pass/fail JSON checkpoint with issues and summary. Use for code review gates on uncommitted/branch diffs.

Git

Bash

OpenAI

code-quality

From conductor

Install

Run in your terminal

npx claudepluginhub ggprompts/my-plugins --plugin conductor

Tool Access

This skill uses the workspace's default tool permissions.

Skill Content

Similar Skills

cache-components

Guides Next.js Cache Components and Partial Prerendering (PPR) with cacheComponents enabled. Implements 'use cache', cacheLife(), cacheTag(), revalidateTag(), static/dynamic optimization, and cache debugging.

cache-components

138.6k

claude-opus-4-5-migration

2 files

Migrates code, prompts, and API calls from Claude Sonnet 4.0/4.5 or Opus 4.1 to Opus 4.5, updating model strings on Anthropic, AWS, GCP, Azure platforms.

claude-opus-4-5-migration

83.2k

evaluation-methodology

1 file

Details PluginEval's skill quality evaluation: 3 layers (static, LLM judge), 10 dimensions, rubrics, formulas, anti-patterns, badges. Use to interpret scores, improve triggering, calibrate thresholds.

plugin-eval

32.9k

Stats

Parent Repo Stars3

Parent Repo Forks0

Last CommitMar 10, 2026

Actions

View Source View Plugin View on GitHub View README

Codex Review Checkpoint

Code review checkpoint that runs codex review on the current worktree.

Note: The gate-runner script runs this automatically for issues with gate:codex-review label. This skill documents the process for manual use.

Quick Start

# Review uncommitted changes
codex review --uncommitted --title "My Review"

# Review branch changes against main
codex review --base main --title "Feature Review"

# Review a specific commit
codex review --commit abc123

What This Skill Does

Determines what to review (uncommitted changes vs branch diff)
Runs codex review (non-interactive, exits on completion)
Parses results into structured format
Writes result to checkpoint file

Workflow

Step 1: Determine Review Scope

# Check for uncommitted changes
git status --porcelain

# Check for branch diff
git diff --quiet main...HEAD && echo "No branch changes" || echo "Has branch changes"

Step 2: Run Codex Review

# For uncommitted changes
codex review --uncommitted --title "Checkpoint Review"

# For branch diff against main
codex review --base main --title "Checkpoint Review"

Codex review:

Runs non-interactively (no terminal interaction needed)
Outputs structured findings with severity levels (P1, P2, P3)
Exits with appropriate status

Step 3: Parse and Write Result

After reviewing, create a structured result:

{
  "checkpoint": "codex-review",
  "timestamp": "2026-01-19T12:00:00Z",
  "passed": true,
  "issues": [],
  "summary": "No critical issues found"
}

Result Fields:

passed: true if no P1/P2 issues found
issues: array of {severity, message, file?, line?}
summary: brief human-readable summary

Step 4: Write Checkpoint File

mkdir -p .checkpoints
cat > .checkpoints/codex-review.json << 'EOF'
{
  "checkpoint": "codex-review",
  "timestamp": "...",
  "passed": true,
  "issues": [],
  "summary": "..."
}
EOF

Decision Criteria

Pass if:

No P1 or P2 issues found
No security vulnerabilities detected
No obvious bugs or logic errors

Fail if:

P1/P2 issues found
Security issues found
Critical bugs detected

Example Output

$ codex review --uncommitted --title "Feature Review"
OpenAI Codex v0.87.0 (research preview)
--------
workdir: /home/user/project
model: gpt-5.2
...

[P2] Guard against missing script before execution — path/to/file.sh:45
  Description of the issue...

[P3] Consider adding error handling — path/to/other.ts:120
  Suggestion details...

Automated Usage

The gate-runner.sh script runs codex review automatically:

./plugins/conductor/scripts/gate-runner.sh --issue ISSUE-ID --worktree /path/to/worktree

This:

Checks issue labels for gate:codex-review
Runs codex review directly (no terminal spawn needed)
Writes checkpoint to .checkpoints/codex-review.json
Returns pass/fail status