ActivityPub MCP

ActivityPub MCP

A comprehensive MCP server to explore and interact with the Fediverse via MCP tools, resources, and prompts.

7
Stars
0
Forks
2
Releases

Overview

ActivityPub MCP Server is a comprehensive MCP server that enables LLMs to explore and interact with the Fediverse through standardized MCP tools, resources, and prompts. It provides a Fediverse Client to talk to ActivityPub servers (Mastodon, Pleroma, Misskey, etc.), WebFinger discovery to locate actors across the network, and a full MCP protocol implementation with read-only Resources, Tools, and Prompts. The server is optimized for LLM interaction patterns and is fully typed in TypeScript with modern ESM. It includes monitoring, health checks, and performance metrics, supporting cross-platform operation on Windows, macOS, and Linux. The project ships with installation scripts, environment configuration, and documentation, and can be tested with the MCP Inspector. Claude Desktop integration is supported via a configuration entry. The architecture emphasizes modular components for WebFinger, remote ActivityPub clients, instance discovery, and data retrieval, enabling LLMs to discover remote actors, timelines, instances, and search across the Fediverse using a complete MCP stack.

Details

Owner
cameronrye
Language
Astro
License
MIT License
Updated
2025-12-07

Features

Resources

Access remote ActivityPub data (actors, timelines, instance info) across fediverse servers.

Tools

Discover and interact with fediverse content through MCP tools.

Prompts

Templates for fediverse exploration and discovery prompts.

Monitoring

Built-in logging and performance metrics to observe MCP server activity.

Health Checks

Server health monitoring and diagnostics to ensure reliability.

Fediverse Client

Interact with existing ActivityPub servers (Mastodon, Pleroma, Misskey, etc.).

WebFinger Support

Resolve actor identifiers across the network using WebFinger.

Security

Built-in security features and input validation to protect data and access.

Audience

LLM developersProvide MCP-enabled interfaces for LLMs to explore and interact with Fediverse data.
AI agentsEnable AI agents to access actors, timelines, and instance information through MCP.
Fediverse researchersSupport researchers in prototyping MCP workflows for data retrieval and discovery.

Tags

FediverseActivityPubMCPWebFingerActor DiscoveryTimelineInstance DiscoveryInstance InfoRemote DataLLMTypeScriptMonitoringSecurity