About butterbase-ai
@butterbase/mcp is the official MCP server for Butterbase, letting MCP clients such as Claude Code, Claude Desktop and Cursor manage schemas, auth, functions, storage, RAG and deploys. It can run locally via npx or be used through a hosted endpoint. It exposes 38 tools across the platform.
Key features
- 01Local npx run or hosted remote endpoint
- 02Schema, migration and row data management
- 03Auth configuration, OAuth and user management
- 04Function deployment, invocation and storage or KV management
- 05RAG content management and querying
- 06Frontend deployment, previews and RLS configuration
Use cases
- 01Managing a Butterbase app's database schema from Claude Code or Cursor
- 02Deploying and invoking functions through an AI client
- 03Configuring auth, storage and RAG retrieval for an app
MCP support
This project is an MCP server that exposes 38 tools to MCP clients for managing Butterbase apps, data, auth, functions and deployments.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://api.butterbase.ai/mcpclaude mcp add --transport http butterbase-ai https://api.butterbase.ai/mcp{"mcpServers":{"butterbase-ai":{"type":"http","url":"https://api.butterbase.ai/mcp"}}}Open in Cursor{"servers":{"butterbase-ai":{"type":"http","url":"https://api.butterbase.ai/mcp"}}}Open in VS Code{"mcpServers":{"butterbase-ai":{"type":"http","url":"https://api.butterbase.ai/mcp"}}}{"mcpServers":{"butterbase-ai":{"serverUrl":"https://api.butterbase.ai/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
https://api.butterbase.ai/mcpAuth requiredTools
Details
From the project README@butterbase/mcp
The official Model Context Protocol server for Butterbase — manage schemas, auth, functions, storage, RAG, realtime, and deploys on Butterbase from any MCP-capable client (Claude Code, Claude Desktop, Cursor, Windsurf, etc.).
Install
npx @butterbase/mcp
Or install globally:
npm install -g @butterbase/mcp
butterbase-mcp
Configure
Add to your MCP client config (.mcp.json, claude_desktop_config.json, etc.):
{
"mcpServers": {
"butterbase": {
"command": "npx",
"args": ["-y", "@butterbase/mcp"],
"env": {
"BUTTERBASE_API_KEY": "bb_sk_your_key_here"
}
}
}
}
Get an API key at butterbase.ai.
Hosted alternative
Prefer to skip the local install? Butterbase also runs a hosted MCP endpoint:
{
"mcpServers": {
"butterbase": {
"url": "https://api.butterbase.ai/mcp",
"headers": {
"Authorization": "Bearer ${BUTTERBASE_API_KEY}"
}
}
}
}
What you can do
The server exposes 38 tools across Butterbase's surface area, including:
- Apps & regions —
init_app,manage_app(actions: list/delete/pause/get_config/update_access_mode/secure/update_cors/set_visibility/preview_clone_env_vars/clone/get_clone_job/find_templates/set_clone_webhook/link_substrate/unlink_substrate/set_substrate_autopropagate/move/move_status/teardown_source_replica/get_env/update_env/publish_template_release/list_template_releases/get_template_release/check_template_updates/update_from_template),list_regions - Schema —
manage_schema,manage_migrations - Auth —
manage_auth_config,manage_auth_users,manage_oauth,manage_people - Data —
select_rows,insert_row,seed_database - Functions —
deploy_function,invoke_function,manage_function - Storage & KV —
manage_storage,manage_kv - Frontends & previews —
create_frontend_deployment,manage_frontend,manage_edge_ssr,manage_preview,promote_preview - RAG —
manage_rag_content,rag_query - Realtime & Durable Objects —
manage_realtime,manage_durable_objects - AI gateway & agents —
manage_ai,manage_agents - RLS —
manage_rls - Repo —
manage_repo - Integrations & billing —
manage_integrations,list_partner_apis,manage_billing - API keys & audit logs —
manage_api_keys,query_audit_logs - Docs, suggestions & hackathons —
butterbase_docs,submit_suggestion,prep_and_submit_hackathon_entry
See the Butterbase docs for the full list.
License
Apache-2.0 © NetGPT Inc.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.