Get a second opinion in Claude Code from GPT, Gemini, and Grok - plus 300+ more models through OpenRouter, including Qwen, Kimi, and DeepSeek. Seven domain experts (Architect, Code Reviewer, Security Analyst, and four more) review your plans, find bugs, and debate edge cases until they agree.
One model is a guess. Three that agree is a plan. → read the blog post
📸 See a full /consensus run: round 1 disagreement to round 5 convergence
... a few moments later ...
📸 See /ask-all stage a 2-round architect debate: three models, three verdicts, then each critiques the others - disagreement matrix included
When three models argue, the real bug reveals itself. Round 1 = independent top findings. Round 2 = each model dunks on the others' picks. The disagreement matrix shows where they diverge; the conclusion shows what to actually fix first.
Claude can ask GPT, Gemini, Grok, or any OpenAI-compatible model (via OpenRouter) for help through MCP. The plugin handles the wiring for each provider so you just write the prompt. Each expert has a distinct specialty and can advise or implement.
You can use any subset of the providers. The plugin detects which are configured and routes
accordingly. OpenRouter is advisory-only and config-driven: models are declared in
~/.config/deliberation/config.json (Windows: %APPDATA%\deliberation\config.json; override
with DELIBERATION_CONFIG) and hot-reload without restarting Claude Code.
| What you get | Why it matters |
|---|---|
| 7 domain experts | The right specialist for each problem type |
| GPT, Gemini, Grok, or OpenRouter models | Use your preferred provider(s) |
| Dual mode | Experts analyze (read-only) or implement (write) |
| Auto-routing | Claude detects when to delegate from your request |
| Synthesized responses | Claude interprets expert output, never raw passthrough |
1. Add the marketplace - antonbabenko/agent-plugins
/plugin marketplace add antonbabenko/agent-plugins
2. Install the plugin
/plugin install deliberation@antonbabenko
3. Run setup
/deliberation:setup
Claude now routes complex tasks to your GPT, Gemini, Grok, and OpenRouter experts (Grok and OpenRouter advise; GPT and Gemini can also implement).
The orchestration server is also published on its own - npm @antonbabenko/deliberation-mcp, Official MCP Registry name io.github.antonbabenko/deliberation.
One-click install:
Manual config for any MCP clients
Add this to your host's MCP config (most hosts use the mcpServers key):
{
"mcpServers": {
"deliberation": {
"command": "npx",
"args": ["-y", "@antonbabenko/deliberation-mcp"],
"env": {
"XAI_API_KEY": "xai-...",
"OPENROUTER_API_KEY": "sk-or-v1-..."
}
}
}
}The env block is how you set provider keys outside Claude Code. GPT and Gemini do not read keys here - they use the codex and agy CLIs (logged in separately), so drop those lines if you only use GPT/Gemini. XAI_API_KEY enables Grok; OPENROUTER_API_KEY enables OpenRouter (which also needs models declared in ~/.config/deliberation/config.json - the canonical XDG path, Windows %APPDATA%\deliberation\config.json - or point elsewhere with DELIBERATION_CONFIG). The one-click buttons above cannot carry secrets - add the env block by hand after installing.
Per-host config location and the key it expects:
| Host | Config | Key |
|---|---|---|
| Claude Code | claude mcp add deliberation -- npx -y @antonbabenko/deliberation-mcp (or project .mcp.json) |
mcpServers |
| Claude Desktop | ~/Library/Application Support/Claude/claude_desktop_config.json (macOS), %APPDATA%\Claude\claude_desktop_config.json (Windows) |
mcpServers |
| Cursor | ~/.cursor/mcp.json (global) or .cursor/mcp.json (project) |
mcpServers |
| VS Code | .vscode/mcp.json - note: each entry needs "type": "stdio" |
servers |
| Codex CLI | ~/.codex/config.toml - TOML, e.g. [mcp_servers.deliberation] |
mcp_servers |
| Gemini CLI | ~/.gemini/settings.json |
mcpServers |
| Windsurf | ~/.codeium/windsurf/mcp_config.json |
mcpServers |
| Zed | settings.json |
context_servers |
| Cline | the extension's MCP settings (Cline panel -> MCP Servers) | mcpServers |
Provider prerequisites are the same as the plugin (see Requirements): the Codex CLI for GPT, agy for Gemini, XAI_API_KEY for Grok, and OPENROUTER_API_KEY plus ~/.config/deliberation/config.json for OpenRouter (Windows: %APPDATA%\deliberation\config.json; override the config path with DELIBERATION_CONFIG).
Tools exposed: ask-all, consensus, ask-gpt / ask-gemini / ask-grok / ask-openrouter, and the seven experts (architect, plan-reviewer, scope-analyst, code-reviewer, security-analyst, researcher, debugger).
The package also ships a deliberation-setup bin. Run it once with npx -y --package @antonbabenko/deliberation-mcp deliberation-setup to write a starter ~/.config/deliberation/config.json (it never overwrites an existing one). The plain npx -y @antonbabenko/deliberation-mcp form runs the default bin (the server), which is what your MCP host launches. For host rule wiring, see AGENTS.md and the per-host snippets in examples/.
You need at least one provider:
- Codex CLI (GPT):
npm install -g @openai/codex, thencodex login. - Antigravity CLI: Getting Started with Antigravity CLI and Migrating from Gemini CLI, then run
agyand login. - 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). - OpenRouter: no CLI; the bridge ships with the plugin (needs Node 18+). Set
OPENROUTER_API_KEY(get a key at https://openrouter.ai/keys), then declare models in~/.config/deliberation/config.json(Windows:%APPDATA%\deliberation\config.json; override withDELIBERATION_CONFIG). Works with any OpenAI-compatible endpoint (Ollama, vLLM, LM Studio, HuggingFace Inference) - auth is skipped automatically when the key env var is empty.
Bundled with the plugin (available once installed):
| Command | Purpose |
|---|---|
/deliberation:setup |
Configure Codex/Gemini/Grok/OpenRouter MCP servers + orchestration rules |
/deliberation:consensus |
🔥🔥🔥 Arbiter-mediated GPT + Gemini + Grok + Claude convergence loop |
/deliberation:ask-all |
🔥 GPT + Gemini + Grok (+ configured OpenRouter models) in parallel, synthesized |
/deliberation:ask-gpt |
One-shot GPT (Codex) second opinion |
/deliberation:ask-gemini |
One-shot Gemini second opinion |
/deliberation:ask-grok |
One-shot Grok (xAI) second opinion (advisory-only) |
/deliberation:ask-openrouter |
One-shot OpenRouter model second opinion (advisory-only) |
/deliberation:uninstall |
Remove MCP config, rules, and aliases |
/deliberation:grok-files |
List, prune, or gc Grok-uploaded files (storage + local cache cleanup) |
/setup can also install short aliases (/ask-gpt, /ask-gemini, /ask-grok, /ask-all, /consensus, /grok-files) into ~/.claude/commands/. This is opt-in. Existing same-named commands are kept by default; setup asks before overwriting any of them. /uninstall removes an alias only if it is byte-identical to the bundled copy.
| Expert | What they do | Example triggers |
|---|---|---|
| Architect | System design, tradeoffs, complex debugging | "How should I structure this?" / "What are the tradeoffs?" |
| Plan Reviewer | Validate plans before you start | "Review this migration plan" / "Is this approach sound?" |
| Scope Analyst | Catch ambiguities early | "What am I missing?" / "Clarify the scope" |
| Code Reviewer | Find bugs, improve quality | "Review this PR" / "What's wrong with this?" |
| Security Analyst | Vulnerabilities, threat modeling | "Is this secure?" / "Harden this endpoint" |
| Researcher | External libraries, docs, best practices | "How do I use X?" / "Find examples of Y" |
| Debugger | Root-cause analysis, minimal fixes | "Why does this crash?" / "Debug this failing test" |
- Architecture decisions - "Should I use Redis or in-memory caching?"
- Stuck debugging - after two or more failed attempts, get a fresh perspective
- Pre-implementation - validate a plan before writing code
- Security concerns - "Is this auth flow safe?"
- Code quality - a second opinion on your implementation
- Simple file operations (Claude handles these directly)
- First attempt at any fix (try yourself first)
- Trivial questions (no need to delegate)
Describe your task. Claude detects when an expert helps and delegates automatically:
You: "Is this authentication flow secure?"
Claude: routes to the Security Analyst, then synthesizes the findings.
You can also ask explicitly: "Ask GPT to review this architecture", "Ask Gemini to...", or "Ask Grok to...". Each expert runs read-only for analysis or with write access to apply fixes, and Claude picks the mode from your request.
Or invoke the slash commands directly - see Commands above.
/ask-gpt, /ask-gemini, /ask-grok, and /ask-all are the quick commands: each dispatches one or three external models, Claude reads the output, and you get one synthesized answer. Single shot, no loop, no peer round.
/consensus is the heavy one. Same parallel dispatch, but with a peer-review round and a multi-round loop that stops only when the models agree. The cost: the orchestrator (Claude) writes the review prompt, casts a vote, decides which objections are real, and runs the loop. Left alone, that setup can quietly rubber-stamp its own plan. Four guards stop that.
See the detailed diagram with bias guards and per-model flow
The four guards:
- Blind verdict. Claude posts its own verdict (APPROVE / REQUEST CHANGES / REJECT) in a message sent before the one that calls the panel. The pre-commitment sits in the transcript, so Claude cannot reshape its opinion after seeing the others.
- Peer review. Stage 2: each external model rates the OTHER models' answers blind, with identity stripped best-effort. Not-viable votes become candidate critical issues for the arbiter to weigh. Pattern adapted from karpathy/llm-council.
- No self-approval. A round converges only when every responding external approves and at least one external actually answered. Claude's own approval never carries a round by itself. A provider that errors (an unconfigured Grok returning
missing-auth, for example) drops out of the count instead of jamming the loop. - No silent dismissal. Every critical issue that gets dismissed or deferred ships with a one-line reason in the final report, including the times Claude walks back one of its own blind objections.
The /ask-* commands carry a lighter version of the same rule. The external model only advises: Claude reads the output, applies its own judgment, and owns the synthesized answer. When the models agree, that is input, not a verdict.
Deep dive: how a single /consensus round actually runs
Each /consensus round runs three stages:
- Stage 1 - parallel verdicts. Claude commits a blind verdict first; GPT, Gemini, and Grok review the plan in parallel and emit APPROVE / REQUEST CHANGES / REJECT plus a list of critical issues.
- Stage 2 - blind cross-review (conditional). Each external reviewer rates the OTHER reviewers' anonymized answers as viable or not-viable, with a one-line reason and a category. Reviewer identity is stripped best-effort (preamble plus self-references). House styles may still leak, so reviewers are instructed to score substance, not style. Not-viable votes become candidate critical issues for Stage 3.
- Stage 3 - arbiter adjudication. Claude reconciles Stage 1 verdicts, Stage 2 candidate issues, and its own blind verdict. For each issue it picks accept, dismiss (with reason), or defer. Then it revises the plan for the next round.
The loop stops when all responding externals approve, zero critical issues remain accepted, and Claude adjudicates APPROVE. Hard cap at 5 rounds.
Stage 2 trigger. Round 1 always. Round 2 onwards, Stage 2 fires only when Stage 1 has divergence OR the previous Stage 2 surfaced an arbiter-accepted not-viable issue (a one-round lookback that catches rubber-stamp convergence after a divergent round).
Stage 2 activation boundary. Stage 2 is skipped on sandbox: workspace-write runs (Claude is writing code, not reviewing prose). Plan reviews that merely contain embedded diff text as prose still run Stage 2.
Cost ceiling. Stage 2 adds at most about 60% to the call count in the worst case (5 rounds, Stage 2 firing every round). Typical convergence (2 to 3 rounds with partial Stage 2 fires) adds 25 to 50%.
Scoring vocabulary and operator-visible debug details live in TECHNICAL.md.
Every expert supports two modes, chosen automatically from your request:
| Mode | Sandbox | Use when |
|---|---|---|
| Advisory | read-only |
Analysis, recommendations, reviews |
| Implementation | workspace-write |
Making changes, fixing issues |
OpenRouter models are declared in ~/.config/deliberation/config.json - the canonical
XDG path (Windows: %APPDATA%\deliberation\config.json). You can override the path
with DELIBERATION_CONFIG. The file is the live single source of
truth: changes to models, routing, or the providers.openrouter block hot-reload
without restarting Claude Code. Toggling a built-in provider (codex / gemini / grok)
still requires /setup.
The config has four sections: providers (transport / connection per provider),
models (named model records keyed by id), routing (fan-out policy), and
consensus.arbiter (who synthesizes the verdict). The $schema key gives editors
validation and autocomplete - VS Code needs no extension.
Minimal example:
{
"$schema": "https://raw.githubusercontent.com/antonbabenko/deliberation/master/config/config.schema.json",
"version": 1,
"providers": {
"codex": { "enabled": true },
"gemini": { "enabled": true },
"grok": { "enabled": true, "apiKeyEnv": "XAI_API_KEY" },
"openrouter": {
"enabled": true,
"apiKeyEnv": "OPENROUTER_API_KEY",
"apiBase": "https://openrouter.ai/api/v1",
"defaultModel": "openai/gpt-4.1-mini",
"defaults": { "reasoningEffort": "medium" }
}
},
"models": {
"gpt-4-or": {
"provider": "openrouter",
"model": "openai/gpt-4.1",
"askAll": true,
"consensus": false
},
"claude-arb": {
"provider": "openrouter",
"model": "anthropic/claude-haiku-4-5",
"askAll": true,
"consensus": true,
"reasoningEffort": "high"
}
},
"routing": { "maxFanout": 3 },
"consensus": { "arbiter": { "model": "claude-arb" } }
}Browse model slugs at openrouter.ai/models;
the model field takes any slug listed there. Each record's provider must be
"openrouter" in v1 (codex / gemini / grok are managed by their own CLI / API).
reasoningEffort (low / medium / high) sets how hard a reasoning model
thinks. Put it on providers.openrouter.defaults to cover every model, or on a single
record to override the default for that one. Precedence runs call argument over
per-record override over defaults.
/ask-all includes records where askAll !== false, capped to routing.maxFanout.
/consensus includes records where consensus === true, with no fanout cap (a warning
is emitted when more than 3 models participate). consensus.arbiter picks who synthesizes:
a shorthand string ("auto" / "host" / "codex" / "gemini" / "grok") or
{ "model": "<id>" } naming a record (even an out-of-panel one). Implementation tasks
always route to Codex or Gemini - never OpenRouter.
For the full schema, the $schema / VS Code validation story, apiBase override matrix
(Ollama, vLLM, LM Studio, HuggingFace), file-attachment caps, session model persistence,
consensus cost model, and error kinds, see
TECHNICAL.md - OpenRouter bridge.
For provider defaults, environment variables, and manual MCP setup, see TECHNICAL.md.
2.0 renames the plugin and drops 1.x compatibility. One-time steps:
- Move your config to the canonical path:
mkdir -p ~/.config/deliberation && mv ~/.claude/claude-delegator/config.json ~/.config/deliberation/config.json(or setDELIBERATION_CONFIGto a custom path;CLAUDE_DELEGATOR_CONFIGis no longer read). - Re-run
/deliberation:setupto register the renamed MCP servers. The old barecodex/gemini/grok/openrouterregistrations are not auto-removed. If you had them, remove each one (the CLI takes one name per call):claude mcp remove --scope user codex claude mcp remove --scope user gemini claude mcp remove --scope user grok claude mcp remove --scope user openrouter - Slash commands moved from
/claude-delegator:*to/deliberation:*(the short aliases/ask-gptetc. are unchanged). - Optional cleanup:
rm -rf ~/.claude/rules/delegator ~/.claude/cache/claude-delegator. - Restart Claude Code.
Maintained by Anton Babenko - LinkedIn, X/Twitter.
Contributions welcome. See CONTRIBUTING.md for the workflow, commit conventions, and the automated release process.
Deliberation started as a fork of jarrodwatts/claude-delegator - credit to Jarrod Watts for the original solution and inspiration. Original work and MIT copyright are retained. This fork adds Grok support, Gemini bridge reliability (timeout and trust recovery), provider configuration overrides, and the bundled delegation commands. It is not an official continuation of the upstream project.
Expert prompts are adapted from oh-my-openagent (snapshot 03eb9fff, 2026-05-25) and PAL MCP server (Apache-2.0, snapshot 7afc7c1, 2026-05-25).