opencode

opencode plugin for Claude Code

Use opencode from inside Claude Code for code reviews or to delegate tasks to opencode.

This plugin is for Claude Code users who want an easy way to start using opencode from the workflow they already have.

Inspired by and structurally modeled on openai/codex-plugin-cc — same slash-command shape, same companion-script architecture, retargeted at opencode's REST server instead of Codex's JSON-RPC app server.

What You Get

• /opencode:review for a normal read-only opencode review
• /opencode:adversarial-review for a steerable challenge review
• /opencode:rescue to hand a free-form task to opencode through a subagent
• /opencode:status, /opencode:result, /opencode:cancel to manage tracked background jobs
• /opencode:broker and /opencode:sessions to control the long-lived opencode serve process and list its sessions
• /opencode:setup to verify the opencode CLI is installed

Requirements

• An opencode-compatible model. opencode supports any provider on models.dev — Anthropic, OpenAI, OpenRouter, Moonshot, GitHub Copilot, etc. The plugin doesn't pick a model for you; whatever opencode is configured to use is what you get.
• Node.js 20 or later (for the companion script that ships with the plugin).

Install

Add the marketplace in Claude Code:

/plugin marketplace add JoelFrancisco/opencode-plugin-cc

Install the plugin:

/plugin install opencode@opencode-plugin-cc

Reload plugins:

/reload-plugins

Then run:

/opencode:setup

/opencode:setup will tell you whether opencode is installed. If it isn't, install it with:

curl -fsSL https://opencode.ai/install | bash

If opencode is installed but not authenticated yet, run:

!opencode auth login

After install, you should see:

• the slash commands listed below
• the opencode:opencode-rescue subagent in /agents

One simple first run:

/opencode:review --background
/opencode:status
/opencode:result

Usage

/opencode:review

Runs a normal opencode review on your current work.

[!NOTE] Code review for multi-file changes might take a while. It's generally recommended to run it in the background.

Use it when you want:

• a review of your current uncommitted changes
• a review of your branch compared to a base branch like main

Supports --base <ref> for branch review, --model <provider/model> to override opencode's configured model, plus --wait and --background to choose execution mode. It is not steerable and does not take custom focus text — use /opencode:adversarial-review for that.

Examples:

/opencode:review
/opencode:review --base main
/opencode:review --background
/opencode:review --model anthropic/claude-sonnet-4-6

This command is read-only. When run in the background you can use /opencode:status to check progress and /opencode:cancel to abort.

/opencode:adversarial-review

Runs a steerable review that questions the chosen implementation and design.

It can be used to pressure-test assumptions, tradeoffs, failure modes, and whether a different approach would have been safer or simpler.

Same scope flags as /opencode:review (--base <ref>, --wait, --background, --model). Unlike /opencode:review, it can take extra focus text after the flags — that becomes the Focus areas: line in the prompt.

Examples:

/opencode:adversarial-review
/opencode:adversarial-review --base main challenge whether this caching design is safe
/opencode:adversarial-review --background look for race conditions and question the chosen approach

This command is read-only. It does not fix code.

/opencode:rescue

Hands a task to opencode through the opencode:opencode-rescue subagent.

Use it when you want opencode to:

• investigate a bug
• try a fix
• continue a previous opencode task in this workspace
• take a faster pass with a smaller model

[!NOTE] These tasks can take a long time depending on the model and the work — it's generally recommended to run them in the background.

Supports --background, --wait, --resume, --fresh, and --model <provider/model>. If you omit --resume and --fresh, the plugin offers to continue the latest rescue thread for this workspace when one exists.

Examples:

/opencode:rescue investigate why the tests started failing
/opencode:rescue fix the failing test with the smallest safe patch
/opencode:rescue --resume apply the top fix from the last run
/opencode:rescue --background investigate the regression
/opencode:rescue --model openrouter/moonshotai/kimi-k2-0905 sketch a faster algorithm

You can also just ask for a task to be delegated: