What is dual-model-mcp CLAUDE.md?

Project instructions for dual-model-mcp, an MCP server that asks multiple language models for parallel second opinions, with optional synthesis into one result. It can run locally or as a remote service.

They are for developing, configuring, deploying, and connecting the dual-model MCP server to local clients or a remote Claude connector.

They provide the project’s architecture, deployment setup, model configuration, and security details needed to maintain it safely.

How Firnschnee/dual-model-mcp configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to Firnschnee/dual-model-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/Firnschnee/dual-model-mcp/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/Firnschnee/dual-model-mcp

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/firnschnee/dual-model-mcp/claude-md/github.svg)](https://agentmods.dev/instructions/firnschnee/dual-model-mcp/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/firnschnee/dual-model-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/firnschnee/dual-model-mcp/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/firnschnee/dual-model-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/firnschnee/dual-model-mcp/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 1,526 This file is loaded in full into every session.

When invoked 1,526 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.01526	$0.01526
Opus 5.5	$0.00610	$0.00610
Sonnet 5.5	$0.00305	$0.00305
Haiku 4.5	$0.00153	$0.00153

Measured 2mo ago against content hash c0c79afbe658, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

dual-model-mcp CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

CLAUDE.md · 92 lines

How it starts

The opening of the file, as written. The whole thing — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.

dual-model-mcp - Arbeitskontext

MCP-Server fuer parallele LLM-Zweitmeinungen ueber OpenRouter, optional mit Synthese. Installation und Nutzung stehen im README; diese Datei ist Kontext fuer die Arbeit am Code, nicht die Nutzeranleitung.

Architektur

src/index.ts - STDIO-Entry (lokal: Claude Code / Desktop / Cherry Studio).
src/http.ts - Streamable-HTTP-Entry (remote: claude.ai Web/Mobile als Custom Connector).
src/server.ts - gemeinsamer Kern: Tool-Definition, OpenRouter-Calls, Synthese. Beide Entries teilen sich diesen Kern.
N Modelle statt fix zwei: MODELS als kommaseparierte Liste in .env, kein Rebuild noetig. Optionale Synthese (synthesize: true, Default-Modell Gemini 3.8 Flash). Native fetch, kein HTTP-Client-Dependency. Node 20+ (AbortSignal.any).
Positionierung seit 2026-09-05: Eskalationsstufe, nicht Alltag. Defaults Fable 5.1 + GPT-6 Astra, reasoning.effort=high (OpenRouter-einheitlicher Parameter, pro Anbieter uebersetzt), MAX_TOKENS 32000, Timeout 600 s. Begruendung: der Aufrufer ist im Alltag Opus 5; ein zweites Mittelklassemodell bringt keine Diversitaet, die den Preis wert waere.
Zweites Tool ask_gpt (seit 1.2.0): ein Modell (GPT-5.6 Sol, effort high, 32000 Tokens), keine Synthese, gleiche Ausgabe. Beide Tools teilen sich runQuery(). Eigenes Tool statt Parameter, damit der Aufrufer (claude.ai waehlt nach Beschreibung) fuer die Alltagsfrage nicht Fable+Astra zieht. Slash-Command /gpt (~/.claude/commands/gpt.md).
Synthese ist bewusst Opt-in und im /dual-Command aus: der Aufrufer liest ohnehin alle Rohantworten, die Synthese spart ihm keine Tokens, und das Synthese-Modell sieht nur den Prompt, nicht den Session-Kontext.
temperature wird nur gesendet, wenn explizit gesetzt: Fable, GPT-6, Sonnet 5 unterstuetzen den Parameter laut OpenRouter-Modellliste nicht, OpenRouter verwirft ihn dann still (verifiziert 2026-09-05, kein Fehler).

Remote-Deployment (Hetzner, geteilt mit co-brain)

Read the full file on GitHub · 92 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

2mo ago Changed · +34 lines · +720 tokens per session c0c79afbe658
2mo ago First seen · 58 lines · 806 tokens per session scan A 3cb5284673dc

Subscribe to this mod's changes

dual-model-mcp CLAUDE.md is an instructions file published in the GitHub repository Firnschnee/dual-model-mcp (4 stars, last pushed 1mo ago), licensed MIT. It adds 1,526 tokens to every session, about $0.0061 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Building agents

dual-model-mcp CLAUDE.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
dual-model-mcp CLAUDE.mdFirnschnee/dual-model-mcp	Instructions file	4	1mo ago	MIT	A	1,526	Claude Code
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	today	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	today	MIT	A	4,680	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	5,516	GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	275k	2d ago	MIT	A	4,313	Codex, OpenCode

dual-model-mcp: Instructions file for Claude Code