By Ryan-M-Frank
Two AI coding assistants in one repo, behind a hard lock. Claude and OpenAI Codex plan, implement, and review each other's work without colliding.
Diagnose the Volley environment in one shot - Codex CLI, the bundled MCP bridge, the platform terminal, and the per-repo lock state. Use when setup misbehaves or before starting a session in a new repo.
Use when a plan is ready to build and you want Codex to implement it. Spawns a visible Windows Terminal tab running Codex with a pre-built prompt that points it at the plan and tells it to log progress to .volley/IMPLEMENTATION-LOG.md. Sets the STATE lock to ACTIVE=codex so Claude won't touch source files until Codex finishes. Returns control to the user with a clear "watch the tab" message.
Use after Codex has implemented a plan. Claude reviews the diff (defaults to uncommitted changes; accepts a git revision range as argument), checks it against PLAN.md and acceptance criteria, and writes a verdict to .volley/CODE-REVIEW.md. Inline review - does not call Codex.
Use to send a plan document to Codex for review via MCP. Defaults to the most recently modified PLAN.md under .planning/, or accept an explicit path argument. Codex's review is written to .volley/PLAN-REVIEW.md and surfaced inline. Fast - completes in seconds.
Use to send a GitHub PR's diff to Codex for review via MCP. Required argument is the PR number. Output written to .volley/PR-REVIEW-<num>.md and shown inline. Asks for explicit user opt-in before posting as a PR comment.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Volley is a small bundle of Claude Code skills that lets OpenAI Codex CLI work alongside Claude in the same repository — planning, implementing, and reviewing each other's work — without either one trampling files the other is editing.
You stay in Claude. Codex shows up when you need it, leaves when it's done, and the two never write to the repo at the same time.
If you've ever:
git rebase --interactive for AI sessions...you've hit the limits of single-assistant flows. Volley adds a second assistant with a hard lock so they collaborate instead of collide.
Plan / PR review — fast, returns in seconds (over MCP)
You + Claude ──/volley:review-plan──▶ Codex reads the plan
You + Claude ◀──────── verdict ─────── and returns a critique
(→ .volley/PLAN-REVIEW.md)
Implementation — long-running, in a visible terminal
You + Claude ──/volley:implement───▶ Codex writes the code in
(hands off, waits on the lock) its own terminal tab and
You + Claude ◀──── STATE released ─── logs progress as it goes
You + Claude ──/volley:review-code──▶ Claude reviews the diff
A .volley/STATE file records which assistant is currently authoritative (ACTIVE=claude or ACTIVE=codex), a PID, and a timestamp. Every skill refuses to act unless the lock matches it. No race, no clobbered edits, no "wait, who wrote this?"
# 1. Install the plugin
/plugin marketplace add Ryan-M-Frank/volley
/plugin install volley@volley
# If Codex tools aren't available yet: /reload-plugins (or restart), then:
/volley:setup
# 2. In a repo where you want to use it
cd my-project
claude # start Claude Code
# 3. Inside Claude Code
/volley:setup # one-time: confirms the bundled Codex MCP, scaffolds .volley/
/volley:status # sanity-check the install
Once installed, the skills below are available in any repo.
| Skill | What it does | Calls Codex via |
|---|---|---|
/volley:setup | One-time: verifies Codex, confirms the bundled MCP is reachable, scaffolds .volley/ | (smoke test) |
/volley:status | Inspect the lock state, lock age, stale-PID detection | (local only) |
/volley:unlock | Force-clear a stuck STATE lock (escape hatch) | (local only) |
/volley:diagnose | Health-check: Codex CLI, MCP reachability, platform terminal, and lock state | (local only) |
/volley:review-plan | Hand a plan document to Codex for review; output to .volley/PLAN-REVIEW.md | MCP (fast) |
/volley:review-pr <num> | Hand a GitHub PR diff to Codex for review; output to .volley/PR-REVIEW-<num>.md | MCP (fast) |
/volley:implement | Hand a plan to Codex; Codex runs in a visible terminal tab while Claude waits | Terminal spawn (long-running) |
/volley:review-code | Claude reviews Codex's diff against the plan; output to .volley/CODE-REVIEW.md | (local; no Codex call) |
Two transports for two latency profiles. Short ops (plan review, PR review) go through MCP and return in seconds. Long ops (implementation) spawn a visible Codex terminal so you can watch progress without blocking Claude's session.
codex login)/volley:implement only:
wt) + Git Bash (cygpath)npx claudepluginhub ryan-m-frank/volley --plugin volleyReliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Access thousands of AI prompts and skills directly in your AI coding assistant. Search prompts, discover skills, save your own, and improve prompts with AI.
Plugin-safe Claude Code distribution of Antigravity Awesome Skills with 1,561 supported skills.