---
title: "codex-skills"
description: "Agent skills for research, App Store readiness, Codex model routing, repository and Swift review, PR feedback triage, and safe branch cleanup."
canonical: https://agentpluginsdirectory.com/plugins/codex-skills
last-updated: 2026-10-02
---

# codex-skills
Agent skills for research, App Store readiness, Codex model routing, repository and Swift review, PR feedback triage, and safe branch cleanup.
- Slug: codex-skills
- Publisher: Cory Parry
- Repository: https://github.com/coryparrry/codex-skills
- Manifest: plugins/codex-skills/plugin.json
- Version: 0.19.0
- License: MIT
- Category (editorial): code-review
- Skills: 8 (appstore-readiness-audit, codex-routing, continue-deep-research, deep-code-review, git-clean-merged-branch, research-repo-technology, swift-code-review, triage-review-comments)
- MCP servers: 0
- Stars: 3
- Repository created: 2026-05-10
- Repository last pushed: 2026-09-05
- Publisher type: User
- Listing: https://agentpluginsdirectory.com/plugins/codex-skills
- Schema: https://agent-plugins.org/schemas/1.0.0/plugin.schema.json

## What codex-skills does, in the publisher's words

> Eight Codex skills for research, App Store readiness, Sol and Luna task routing, repository and Swift review, PR feedback triage, and safe branch cleanup.

Each skill solves one repeated workflow problem. Its SKILL.md contains the main instructions.

Supporting details stay in references, scripts, or tests.

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/coryparrry/codex-skills/HEAD/README.md

## Skills

- appstore-readiness-audit: Audit Apple apps before App Store upload or review. Use when checking a release candidate for build, runtime, privacy, SDK, account, payment, metadata, compliance, or reviewer-access rejection risks. Return read-only, evidence-backed upload and submission verdicts.
- codex-routing: Use Sol as the advisor and coordinator for coding work, delegating investigation and implementation to Luna subagents.
- continue-deep-research: Continue existing technical or general research from notes, files, URLs, papers, task IDs, or prior reports. Use when verifying or updating findings, investigating open questions, resolving contradictions, or producing a primary-source-cited delta without restarting or losing provenance.
- deep-code-review: Review a repository snapshot or code change for evidence-backed correctness, security, compatibility, and regression risks. Use for deep, exhaustive, repository-wide, adversarial, or AI-generated code review.
- git-clean-merged-branch: Safely clean up a local Git branch after it has been merged on GitHub. Use when the user says short commands like "clean merged branch", "cleanup merged branch", "sort git", "clean up git", "branch merged", "merged branch cleanup", or invokes "git-clean-merged-branch"; also use for longer requests…
- research-repo-technology: Research technologies an existing repository should adopt, adapt, build, or reject. Use when assessing architecture or product opportunities, comparing external projects at source level, or producing evidence-backed recommendations and proofs of concept from the live codebase.
- swift-code-review: Review Swift and Apple-platform changes, including SwiftUI/AppKit behavior, concurrency, persistence, packages, Xcode project/build configuration, mixed-language boundaries, generated code, and unsafe interoperability, for reachable correctness and regression risks. Use directly for a focused Swift…
- triage-review-comments: Use when triaging PR review comments, GitHub review threads, CodeRabbit/Cursor/bot findings, stale or duplicate feedback, resolved-thread candidates, deferred follow-ups, or prevention tests/checks.

Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.