Skip to main content
Glama
unmerged-approaches-mcp

unmerged-approaches-mcp

by samihalawa

unmerged-approaches-mcp

An MCP server that puts one identical brief to several LLMs in parallel and hands back every answer verbatim — unmerged, unranked, unsynthesized.

Most multi-model MCP servers end in a judge, a synthesizer, or a consensus loop. This one deliberately does not. The moment N answers are averaged into one confident paragraph, the single most useful piece of information is destroyed: whether the models actually agreed.

  • Four models independently reaching the same conclusion is corroboration.

  • Two against two is a real open question your one-model answer would have hidden from you.

A synthesizer renders both of those as the same smooth prose. This server refuses to, by design.

What it does

  • Fans one brief out to up to 12 OpenRouter models concurrently — 400+ models, one API key.

  • Returns per-model: the verbatim answer, finish_reason, latency, token usage, cost, and a structured error when a model fails.

  • A slow or failing model becomes one failed entry; it never sinks the panel.

  • Persists bounded, redacted, correlation-ID-addressable debug events.

  • Never merges, ranks, scores, votes on, or picks a winner among the answers.

Related MCP server: Legion MCP Server

Install

stdio (Claude Desktop, Claude Code, Cursor, Cline, Zed)

{
  "mcpServers": {
    "unmerged-approaches": {
      "command": "npx",
      "args": ["-y", "unmerged-approaches-mcp@latest"],
      "env": { "OPENROUTER_API_KEY": "sk-or-v1-..." }
    }
  }
}
claude mcp add unmerged-approaches npx -y unmerged-approaches-mcp

Remote Streamable HTTP

OPENROUTER_API_KEY=sk-or-v1-... UNMERGED_AUTH_TOKEN=your-token npx unmerged-approaches-mcp-http
# or: node src/http.js

Mounts MCP at /mcp and a separate liveness probe at /health.

{
  "mcpServers": {
    "unmerged-approaches": {
      "type": "http",
      "url": "https://your-host.example/mcp"
    }
  }
}

Send the token as Authorization: Bearer <token> or x-api-key. Keep secrets in your client's secret storage, never in a committed config file.

Tools

Tool

Category

What it does

consult

PANEL

One brief → N models in parallel → N verbatim answers, unmerged

list_models

SYSTEM

Live OpenRouter catalog: substring search, free-only filter, cursor pagination

get_status

SYSTEM

Server version, credential validity, remaining credit, default panel, limits

get_debug_logs

DEBUG

Newest-first structured events, filterable by correlationId / level / event / operation / time

All four are read-only and non-destructive. The server exposes no arbitrary HTTP, shell, or SQL primitive.

consult

Argument

Type

Notes

brief

string, required

Sent byte-for-byte identically to every model. Max 200,000 chars.

models

string[]

Exact OpenRouter ids, max 12, deduplicated. Omit for the default panel.

system

string

Optional system prompt, identical for every model.

max_tokens

int

Default 8000, ceiling 32000.

temperature

number

0–2, default 0.7.

timeout_ms

int

Default 180000, max 600000. Per model, not for the batch.

Returns correlationId, requestedModels, answeredCount, completeCount, failedCount, totalCostUsd, wallClockMs, and answers[] with { model, actualModel, ok, complete, finishReason, content, error, latencyMs, usage }.

If every model fails, the call returns isError: true with code ALL_MODELS_FAILED. A partial panel is a success.

Writing a brief that is worth the money

The output is only as good as the isolation of the input. A brief that leaks your preferred answer gets it echoed back by every model, and you will mistake that echo for agreement.

  • Include the goal, hard constraints, what is already decided, what was rejected and why, and what is genuinely unknown.

  • Mark unverified claims as unverified. Distinguish what you observed from what you assumed.

  • Do not name your preferred option, and do not tell the panel what an earlier model said.

  • Ask for independent judgment, not novelty.

Environment variables

Name

Required

Purpose

OPENROUTER_API_KEY

yes

OpenRouter credential. Server-side only; never accepted as a tool argument.

UNMERGED_DEFAULT_MODELS

no

Comma-separated default panel.

UNMERGED_MAX_MODELS

no

Panel size cap. Default 12.

UNMERGED_MAX_TOKENS

no

Default output ceiling per model. Default 8000.

UNMERGED_TIMEOUT_MS

no

Default per-model timeout. Default 180000.

UNMERGED_LOG_CAPACITY

no

Retained debug events. Default 500.

UNMERGED_AUTH_TOKEN

http only

Bearer / x-api-key token for the HTTP transport.

UNMERGED_ALLOWED_ORIGINS

http only

Comma-separated browser origin allowlist. Browsers are refused unless listed.

PORT, HOST

http only

Defaults 8787, 127.0.0.1.

No value is ever written to logs, tool results, or client configuration.

Cost

You pay per model, per call. A 5-model panel is roughly 5× a single call. list_models with free_only: true finds zero-price models; free endpoints time out and rate-limit far more often, which shows up honestly in failedCount.

Security

  • Credentials are read only from the server environment and are never accepted as tool arguments.

  • Anything credential-shaped is stripped from errors and debug events before persistence.

  • The HTTP transport validates Origin, requires a bearer token when one is configured, and compares it in constant time.

  • Returned answers are untrusted third-party model output. If an agent calls this on attacker-influenced input, treat the answers as data to weigh, never as instructions to execute. This is a prompt-injection surface like any MCP tool that returns external content.

Development

npm install
npm run build   # syntax check every entrypoint
npm test        # 9 contract tests incl. a live stdio initialize/list/call handshake

License

MIT

Available Tools

4 tools
consultConsult an unmerged panel of modelsA
Read-only

Send ONE identical brief to several LLMs in parallel and get every answer back verbatim, attributed, unranked and unsynthesized. Use when a decision is consequential or hard to reverse, when you want your own reasoning stress-tested by models that did not see it, or when the user asks for other opinions or approaches. The value is the spread: independent agreement is corroboration, a split is a real open question that a single answer would have hidden from you. This server never merges, ranks or picks a winner.

Inputs: brief (required), models, system, max_tokens, +2 more.

get_debug_logsRead persisted structured debug eventsA
Read-onlyIdempotent

Newest-first structured events for dispatches, per-model failures and validation errors, filterable by correlationId, level, event, operation and time, with cursor pagination. Every consult result carries a correlationId: pass it here to see exactly what happened per model. Credentials are redacted before persistence.

Inputs: event, level, limit, since, +3 more.

get_statusServer, credential and limit statusA
Read-onlyIdempotent

Report server version, whether OPENROUTER_API_KEY is configured and accepted, remaining credit, the default panel and the configured limits. Call this first when consult fails with an authentication error. Never returns the key itself.

Inputs: no named parameters.

list_modelsList available OpenRouter modelsA
Read-onlyIdempotent

Browse the live OpenRouter model catalog to pick an exact panel for consult. Supports substring search, a free-only filter and cursor pagination. Model ids change over time, so resolve them here rather than guessing.

Inputs: sort, limit, query, cursor, +1 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.6/5.0

    Scored across 4 tools

    Disambiguation5/5

    Each tool has a clearly distinct role: consult is the core action, list_models handles model selection, get_status handles configuration/health checks, and get_debug_logs is for diagnostics. There is no meaningful overlap or possible confusion between the tool boundaries.

    Naming Consistency4/5

    get_status, list_models, and get_debug_logs all follow a clean lower_snake verb_noun pattern. consult is a bare verb without an explicit object, which is a minor deviation, but it still reads naturally and fits the overall imperative style.

    Tool Count5/5

    Four tools is well-scoped for a narrow server: one primary operation plus three supporting utilities for configuration, model discovery, and debugging. Every tool earns its place and there is no redundancy.

    Completeness5/5

    The domain is parallel unmerged consultation, and consult fully covers the core workflow while list_models, get_status, and get_debug_logs cover the supporting needs around model selection, auth/config validation, and failure investigation. There are no obvious dead ends or missing operations agents would need.

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      D
      maintenance
      Enables querying multiple AI models in parallel (Claude, Gemini, O3) and synthesizing their responses using anonymous analysis to reduce bias, providing a comprehensive answer.
      1
      29 npm
      193
      MIT
    • F
      license
      Not graded
      quality
      B
      maintenance
      Exposes multiple LLMs as individual tools via the OpenAI Responses API wire format, enabling the calling AI to get second opinions and orchestrate multi-model discussions with a quorum tool.
      137 npm
      1
      -
    • A
      license
      Not graded
      quality
      B
      maintenance
      Fans out queries to multiple independent free LLMs from different providers and returns their answers side by side, enabling sanity-checking of decisions or getting multiple perspectives during coding sessions.
      GPL 3.0