Install
Pick your host and follow its steps. Then run init once and check the result. You need at least one provider; Requirements below lists what each one needs.
First run, on any host
npx -y @antonbabenko/deliberation-mcp initIt finds the keys and CLI logins you already have, writes the config, pings each provider and runs one short demo consensus. Run it again any time: a second run changes nothing and says so. In Claude Code, /deliberation:setup runs it for you. The quickstart has every step.
Check that it works
Run /deliberation:doctor in Claude Code, or call the doctor tool on any other host. It is read-only: it checks the config, provider CLIs and keys, sessions and path drift, and suggests a fix for each problem it finds.
Requirements
Node.js 18 or newer runs the plugin, the bridges and the npm server.
You need at least one provider:
- Codex CLI (GPT):
npm install -g @openai/codex, thencodex login(ChatGPT subscription;codex login --device-authon a remote machine, one login per machine). A ChatGPT Business/Enterprise workspace can setCODEX_ACCESS_TOKENinstead. With neither, codex usesCODEX_API_KEY;OPENAI_API_KEYis never used. - Gemini (API key, the default): no CLI to install. Set
GEMINI_API_KEYto a Google AI Studio key (get one at https://aistudio.google.com/apikey (external)). Advisory calls go to the Gemini API, modelproviders.gemini.apiModel(defaultgemini-flash-latest). - Antigravity CLI (
agy, opt-in): needed only for Gemini implementation (workspace-write) calls, or if you setproviders.gemini.transportto"agy"; with no key set, Gemini also falls back to it. Getting Started with Antigravity CLI (external), then runagyand log in. Read the Antigravity terms (external) first: section 6 lists "Using third party software, tools, or services to access the Service (e.g. using OpenClaw with Antigravity OAuth)" as a breach that "may be grounds for suspension or termination of your Antigravity and/or Gemini CLI accounts". Deliberation drivesagyas third-party software. - Grok (xAI): no CLI to install; the bridge ships with the plugin (needs Node 18+). Set
XAI_API_KEY(get a key at https://console.x.ai (external)). - OpenRouter: no CLI; the bridge ships with the plugin (needs Node 18+). Set
OPENROUTER_API_KEY(get a key at https://openrouter.ai/keys (external)). With no OpenRouter model records, a built-in panel ofz-ai/glm-5.3-flashx(default),openai/gpt-6.1-solandgoogle/gemini-3.8-flashjoinsask-allandconsensus. To pick your own models, declare them in~/.config/deliberation/config.json(Windows:%APPDATA%\deliberation\config.json; override withDELIBERATION_CONFIG); they replace the built-in panel.providers.openrouter.enabled: falseturns OpenRouter off. Works with other OpenAI-compatible endpoints (vLLM, HuggingFace Inference) throughproviders.openrouter.apiBase- auth is skipped automatically when the key env var is empty. Supports custom OpenRouter presets (e.g.cheapest-first) to optimize inference cost and provider routing. - Ollama and LM Studio: keyless by default (set
apiKeyEnvif your server needs a key). Each has its own provider block,providers.ollama(defaulthttp://localhost:11434/v1) andproviders.lmstudio(defaulthttp://localhost:1234/v1); declare models withprovider: "ollama"orprovider: "lmstudio". See SETUP.md.
Windows note (CLI providers). npm install -g on Windows installs a codex.cmd shim, not an
executable, and Node cannot launch a .cmd directly - so GPT calls used to fail with nothing but a
spawn error. Deliberation now resolves the CLI itself (PATHEXT, falling back to the package's own
entry point) and keeps it out of a shell. If resolution ever misses on your machine, point
CODEX_BIN (GPT) or AGY_BIN (Gemini) at the real executable. Grok and OpenRouter are unaffected -
they are HTTP bridges and spawn nothing.
Claude Code on the web note. The web host caps every MCP tool call at 60s
(MCP_TOOL_TIMEOUT=60000). The plugin manifest overrides that per server: each Deliberation
server declares "timeout": 1800000 (30 min, which Claude Code applies ahead of
MCP_TOOL_TIMEOUT) and mirrors it into the server's env so the plugin's own clamp tracks
the real cap. Nothing to set in the environment. A timeout result whose message still names
MCP_TOOL_TIMEOUT=60000 means an older plugin install: claude plugin update deliberation@antonbabenko and start a new session. /deliberation:doctor reports the cap,
the codex credential, and a missing agy (set GEMINI_API_KEY and Gemini needs no agy for advisory calls); the panel skips providers that cannot answer
(panel.unavailable) instead of waiting on them. For GPT, do not copy your laptop's
~/.codex/auth.json into the container: a ChatGPT login's refresh token works once, so both
the other copy stops working as soon as one of them refreshes. On Plus/Pro, the first GPT call
in a session starts codex login --device-auth and shows its link and code (a dialog when the
host supports it, else in the result); approve it and GPT answers. On Business/Enterprise, set
CODEX_ACCESS_TOKEN. See SETUP.md.