phone-a-friend-mcp-server
Enables AI-to-AI consultation for critical thinking and complex reasoning via OpenRouter, allowing one AI to delegate tasks to another AI model.
README
Phone-a-Friend MCP Server ๐ง ๐
An AI-to-AI consultation system that enables one AI to "phone a friend" (another AI) for critical thinking, long context reasoning, and complex problem solving via OpenRouter.
The Problem ๐ค
Sometimes an AI encounters complex problems that require:
- Deep critical thinking beyond immediate capabilities
- Long context reasoning with extensive information
- Multi-step analysis that benefits from external perspective
- Specialized expertise from different AI models
The Solution ๏ฟฝ
Phone-a-Friend MCP Server creates a two-step consultation process:
- Context + Reasoning: Package all relevant context and send to external AI for deep analysis
- Extract Actionable Insights: Process the reasoning response into usable format for the primary AI
This enables AI systems to leverage other AI models as "consultants" for complex reasoning tasks.
Architecture ๐๏ธ
Primary AI โ Phone-a-Friend MCP โ OpenRouter โ External AI (O3, Claude, etc.) โ Processed Response โ Primary AI
Sequential Workflow:
analyze_context- Gather and structure all relevant contextget_critical_thinking- Send context to external AI via OpenRouter for reasoningextract_actionable_insights- Process response into actionable format
When to Use ๐ฏ
Ideal for:
- Complex multi-step problems requiring deep analysis
- Situations needing long context reasoning (>100k tokens)
- Cross-domain expertise consultation
- Critical decision-making with high stakes
- Problems requiring multiple perspectives
Quick Start โก
Configure your MCP client (e.g., Claude Desktop) using the JSON block belowโno cloning or manual installation required.
The uv runner will automatically download and execute the server package if it isn't present.
Add the following JSON configuration to your MCP client and replace <YOUR_API_KEY> with your key:
{
"mcpServers": {
"phone-a-friend": {
"command": "uvx",
"args": [
"phone-a-friend-mcp-server",
"--provider", "openai",
"--api-key", "<YOUR_API_KEY>"
]
}
}
}
That's it! You can now use the
phone_a_friendtool in any compatible client. For more options, see the Advanced Configuration section.
Available Tools ๐ ๏ธ
phone_a_friend
๐ Consult external AI for critical thinking and complex reasoning. Makes API calls to get responses.
fax_a_friend
๐ Generate master prompt file for manual AI consultation. Creates file for copy-paste workflow.
Parameters
phone_a_friend
all_related_context(required): General, non-code context such as constraints, tracebacks, or high-level requirements.file_list(required): Array of file paths or glob patterns. Just pass the paths โ the server automatically reads those files (skips anything in.gitignoreor non-text/binary) and builds the full code context for the external AI.task(required): A clear, specific description of what you want the external AI to do.
fax_a_friend
all_related_context(required): Same as above.file_list(required): Same as above.task(required): Same as above.output_directory(required): Directory where the generatedfax_a_friend.mdmaster prompt file will be saved.
Advanced Configuration ๐ง
This section covers all configuration options, including environment variables, CLI flags, and model selection.
Providers and API Keys
The server can be configured via CLI flags or environment variables.
| Provider | CLI Flag | Environment Variable |
|---|---|---|
| OpenAI | --provider openai |
OPENAI_API_KEY |
| OpenRouter | --provider openrouter |
OPENROUTER_API_KEY |
| Anthropic | --provider anthropic |
ANTHROPIC_API_KEY |
--provider google |
GOOGLE_API_KEY |
CLI Example:
phone-a-friend-mcp-server --provider openai --api-key "sk-..."
Environment Variable Example:
export OPENAI_API_KEY="sk-..."
phone-a-friend-mcp-server
Model Selection
You can override the default model for each provider.
| Provider | Default Model |
|---|---|
| OpenAI | o3 |
| Anthropic | Claude 4 Opus |
Gemini 2.5 Pro Preview 05-06 |
|
| OpenRouter | anthropic/claude-4-opus |
Override with CLI:
phone-a-friend-mcp-server --model "o3"
Override with Environment Variable:
export PHONE_A_FRIEND_MODEL="o3"
Additional Options
| Feature | CLI Flag | Environment Variable | Default |
|---|---|---|---|
| Temperature | --temperature 0.5 |
PHONE_A_FRIEND_TEMPERATURE |
0.4 |
| Base URL | --base-url ... |
PHONE_A_FRIEND_BASE_URL |
Provider default |
Use Cases ๐ฏ
- In-depth Reasoning for Vibe Coding
- For complex algorithms, data structures, or mathematical computations
- Frontend Development with React, Vue, CSS, or modern frontend frameworks
License ๐
MIT License - see LICENSE file for details.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.