AnythingLLM Control Plane MCP Server
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@AnythingLLM Control Plane MCP ServerList all my workspaces."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
module_type: mcp-server
status: active
protocol: mcp
primary_capability: AnythingLLM administration and workspace management
requires: AnythingLLM instance
works_with: AnythingLLM, Antigravity Agent Ecosystem
last_verified: 2026-08-21AnythingLLM Control Plane MCP Server
A Model Context Protocol (MCP) server for managing and controlling AnythingLLM instances via its Admin REST API.
Status and last verified date
Status: Active
Last Verified: 2026-08-21
Related MCP server: anythingllm-mcp
What it does / does not do
What it does:
Manages AnythingLLM workspaces (create, list, delete).
Fetches system environment configuration from AnythingLLM.
Retrieves system-wide or workspace-specific vector counts.
Embeds a powerful ETL pipeline to ingest agent session logs into AnythingLLM (extracting critical interactions, preventing duplicates, chunking context).
What it does not do:
It does not directly manage LLM providers or API keys within AnythingLLM beyond the scope of workspace management.
It does not replace the AnythingLLM web interface; it augments it for programmatic Agent access.
It is not a vector database itself, it interfaces with AnythingLLM's existing storage.
Why an agent would use it
Agents in the Antigravity ecosystem use this server to dynamically create sandboxed workspaces for new tasks, ingest their own conversational session logs as vector memory, and perform cleanup after task completion. This enables agents to build a self-organizing long-term memory system without manual user intervention.
Architecture and dependencies
Language: Python 3.10+
Protocol: MCP (Model Context Protocol) via
mcp/fastmcpDependencies:
mcp>=1.0.0,httpx>=0.27.0Architecture: Provides an interface over HTTP to the AnythingLLM Admin REST API. Includes a sync daemon (
daemon.py,chunker.py,deduplicator.py,parser_antigravity.py) for processing local agent logs and pushing them into AnythingLLM.
Compatibility
Compatible with AnythingLLM instances running the v1 REST API. Built primarily for the Antigravity Agent Ecosystem but conforms to the standard MCP specification.
Quick start and health check
Install dependencies:
pip install -e .Run server directly (health check):
python3 -m anythingllm_mcp_control.serverIf the server starts successfully, it will wait for MCP connections on stdio.
Configuration and environment variables
ANYTHINGLLM_BASE_URL: Base URL of the AnythingLLM API (Default:http://localhost:3001/api/v1). It will automatically append/api/v1if only a host/port is provided.ANYTHINGLLM_API_KEY: Developer API Key for AnythingLLM authentication (required if authentication is enabled on the instance).
Complete MCP tool/API table with side effects
Tool Name | Arguments | Description | Side Effects |
| None | List all workspaces in AnythingLLM | None |
|
| Create a new workspace in AnythingLLM | Creates a new workspace in the AnythingLLM database |
|
| Delete a workspace in AnythingLLM by slug | Permanently deletes a workspace and its associated data in AnythingLLM |
| None | Retrieve system environment and configuration dump | None |
|
| Get vector count (system-wide or workspace-specific) | None |
Security model and trust boundaries
Authentication: Relies on the
ANYTHINGLLM_API_KEYto authenticate against the AnythingLLM instance.Trust Boundaries: The MCP server inherently trusts the host environment for providing correct API keys and URLs. Agents connected to this MCP server can perform destructive actions (
delete_workspace), so access to this server should be restricted to trusted agents or run in a sandbox.Secret Management: Secrets are passed via environment variables and never logged or exposed in responses.
Tests and exact commands
Run the test suite using pytest:
pytest tests/Operations, logs, backup/restore, rollback
Operations: Runs as an MCP server via stdio, managed by the host MCP client.
Logs: Errors related to API interactions are returned as JSON errors directly to the MCP client.
Backup/Restore: Handled entirely by AnythingLLM itself (backing up its internal SQLite/VectorDB). This control plane is stateless.
Rollback: Downgrade the Python package version and restart the MCP client process.
Generic MCP-client example
{
"mcpServers": {
"anythingllm-control": {
"command": "python3",
"args": ["-m", "anythingllm_mcp_control.server"],
"env": {
"ANYTHINGLLM_BASE_URL": "http://localhost:3001/api/v1",
"ANYTHINGLLM_API_KEY": "YOUR_API_KEY_HERE"
}
}
}
}Limitations and roadmap
Limitations: Only supports basic workspace management and vector ingestion. Doesn't support full document management or granular user permissions via MCP yet.
Roadmap: Add document upload and management tools, user management, and enhanced vector search capabilities directly via MCP.
Related TheNovaNodes modules
Antigravity Agent Ecosystem
Antigravity CLI
License
Please refer to the repository license documentation.
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseBqualityDmaintenanceEnables seamless integration with AnythingLLM instances, providing complete workspace management, chat operations, document handling, user administration, and AI agent configuration through natural language.386815MIT
- FlicenseCqualityDmaintenanceMCP server for AnythingLLM that enables AI document chat platform interaction with tools for workspaces, chat, documents, threads, and system operations.17
- FlicenseNot gradedqualityDmaintenanceEnables file system operations, web scraping, and AI-powered search through MCP tools for use by LLM agents.1
- AlicenseNot gradedqualityCmaintenanceMCP server that provides tools to interact with the LiteLLM proxy API, enabling LLM completions, embeddings, image generation, and admin operations.18MIT
Related MCP Connectors
Manage SRG+ hubs, channels, content, assets, users, and workspaces from any MCP-aware AI agent.
Create and manage Lindo AI websites, pages, blogs, clients, and credits from any MCP client.
OCR, transcription, file extraction, and image generation for AI agents via MCP.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/TheNovaNodes/anythingllm-mcp-control'
If you have feedback or need assistance with the MCP directory API, please join our Discord server