Code review with OpenCode
OpenCode, Anomaly's coding agent, in CodePanel's review loop: it connects through MCP and skills, reads AGENTS.md and answers every finding.
OpenCode is Anomaly's coding agent, in the terminal, a desktop app, VS Code, Cursor and other editors with a terminal and a web interface it serves itself. With CodePanel it is the author on the other side of the review: after it pushes, it reads the panel's findings on that commit over MCP, fixes or disputes each one with evidence, and rates every finding. It reads AGENTS.md and CLAUDE.md (only when there is no AGENTS.md), and the setup skill writes the loop into AGENTS.md or CLAUDE.md.
How OpenCode connects
OpenCode connects through an MCP server entry in opencode.json and the codepanel and codepanel-setup skills. The steps are in Other MCP clients.
OpenCode features for the review loop
| Feature | What the docs say | Source |
|---|---|---|
| Maker | Anomaly | opencode.ai |
| Runs in | the terminal, a desktop app, VS Code, Cursor and other editors with a terminal and a web interface it serves itself | opencode.ai, page 2, page 3 |
| MCP | Local and remote servers, with OAuth, under mcp in opencode.json (mcp.servers in OpenCode 2), or added with opencode mcp add |
opencode.ai, page 2, page 3 |
| Hooks | JavaScript or TypeScript plugins that subscribe to events, session.idle among them for a session gone idle |
opencode.ai |
| GitHub Actions | anomalyco/opencode/github, which runs OpenCode in a workflow when a comment mentions /opencode; opencode github install sets it up |
opencode.ai |
| Headless mode | opencode run "<prompt>", with --auto to approve permissions not explicitly denied |
opencode.ai |
| Rules files | AGENTS.md, CLAUDE.md (only when there is no AGENTS.md) |
opencode.ai |
| Models | Over 75 providers through the AI SDK and Models.dev, local models, and OpenCode Zen, models its team has tested. Prices: OpenAI, Anthropic, Google, xAI, Moonshot, Z.ai, DeepSeek, Qwen. | opencode.ai |
Each row was checked on 2026-10-17 against the maker's docs it cites. A feature with no row is one those docs don't describe.
Read next
AGENTS.md for an agent-to-agent review loop
What to put in AGENTS.md so a coding agent answers AI code review on its own: how each agent loads the file, the review section, and one file for every agent.
CLAUDE.md for an agent-to-agent review loop
What to put in CLAUDE.md so Claude Code sees AI code review through to merge: what loads when, where the loop goes, the size target, and sharing with AGENTS.md.