home / mcp / mcp pickaxe server

MCP Pickaxe Server

An MCP (Model Context Protocol) server that connects AI assistants like Claude to the [Pickaxe](https://pickaxe.co) platform. Manage your AI agents, knowledge bases, users, and analytics directly through natural language.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "aplaceforallmystuff-mcp-pickaxe": {
      "command": "node",
      "args": [
        "/path/to/mcp-pickaxe/dist/index.js"
      ],
      "env": {
        "PICKAXE_STUDIO_DEV": "studio-zzz-zzz-zzz",
        "PICKAXE_STUDIO_MAIN": "studio-your-api-key-here",
        "PICKAXE_DEFAULT_STUDIO": "PRODUCTION",
        "PICKAXE_STUDIO_STAGING": "studio-yyy-yyy-yyy",
        "PICKAXE_STUDIO_PRODUCTION": "studio-xxx-xxx-xxx"
      }
    }
  }
}

You set up an MCP (Model Context Protocol) server that lets AI assistants connect to the Pickaxe platform. It enables you to manage agents, knowledge bases, users, and analytics directly from natural language within a single, streamlined workflow.

How to use

Interact with Pickaxe through natural language to perform common MCP tasks. You can fetch and analyze chat histories, manage documents and knowledge bases, handle user accounts and access, list products, and review or switch between multiple Pickaxe studios. Use plain language requests like: create or connect documents to agents, list users and their usage, or compare documents across studios. The server exposes operations that map to these capabilities so you can drive admin and governance tasks from your assistant.

How to install

Prerequisites: you need Node.js 18+ installed on your system.

# Option 1: Install from npm (recommended)

npx mcp-pickaxe

# Or install globally
npm install -g mcp-pickaxe

# Option 2: Clone and build

git clone https://github.com/aplaceforallmystuff/mcp-pickaxe.git
cd mcp-pickaxe
npm install
npm run build

Configuration

Get your Pickaxe Studio API key and configure your MCP client to connect. The server runs as a local process and connects to Pickaxe Studio using the provided API key.

{
  "mcpServers": {
    "pickaxe": {
      "command": "node",
      "args": ["/path/to/mcp-pickaxe/dist/index.js"],
      "env": {
        "PICKAXE_STUDIO_MAIN": "studio-your-api-key-here"
      }
    }
  }
}

Multi-studio configuration

If you work with multiple Pickaxe studios, define separate environment variables for each studio and choose the default studio to use when none is specified.

{
  "env": {
    "PICKAXE_STUDIO_PRODUCTION": "studio-xxx-xxx-xxx",
    "PICKAXE_STUDIO_STAGING": "studio-yyy-yyy-yyy",
    "PICKAXE_STUDIO_DEV": "studio-zzz-zzz-zzz",
    "PICKAXE_DEFAULT_STUDIO": "PRODUCTION"
  }
}

Use cases you can run with this server

Real-world workflows let you monitor security, maintain knowledge bases, audit memory for personalization, and operate across studios without switching keys. These patterns show how you can automate admin tasks and keep data synchronized across your Pickaxe environment.

Troubleshooting

If you encounter issues, check common problems like missing studio configurations or API key permissions, then verify that your environment variables are set correctly for your chosen studio.

Security notes

Keep your Studio API keys secure. Treat them like passwords and rotate them periodically. When using multi-studio setups, ensure each studio key has appropriate access limits.

Notes on operation

The server is designed to be started as a local process that runs the MCP logic and communicates with Pickaxe Studio. When you launch it, you can drive all supported tools from your assistant through natural language.

Available tools

studios_list

List all configured studios and show the current default; used to switch context across studios.

chat_history

Fetch conversation history for a given agent; supports paging and formatting options.

doc_create

Create a document from content or a URL; supports linking to agents.

doc_list

List all documents with pagination to manage your knowledge base.

doc_get

Retrieve a specific document by identifier.

doc_delete

Delete a document from the knowledge base.

doc_connect

Link a document to an agent to enrich its knowledge.

doc_disconnect

Unlink a document from an agent.

user_list

List all users with access and usage stats.

user_get

Get details for a specific user by email.

user_create

Create a new user with initial access and products.

user_update

Update user details, product access, or usage.

user_delete

Remove a user from the system.

user_invite

Send invitations to new users via email.

products_list

List available products and bundles to assign access.

memory_list

List defined memory schemas for personalization data.

memory_get_user

Retrieve memories collected for a specific user.