Bluesky MCP Server

The Bluesky MCP Server bridges Claude and the AT‑Protocol, allowing an AI assistant to act as a fully authenticated Bluesky user. By exposing a rich set of tools—authentication, posting, interaction, and timeline retrieval—the server eliminates the need for developers to write custom HTTP clients or handle OAuth flows. Instead, they can simply call high‑level, typed tools that encapsulate the underlying protocol logic.

For developers building conversational agents or content automation workflows, this server solves a two‑fold problem: first, it abstracts the complexity of Bluesky’s decentralized API; second, it gives AI agents genuine publishing capabilities. Whether you want a chatbot to post daily insights, schedule announcements, or engage with community content on your behalf, the MCP server provides a secure, credential‑managed interface that respects Bluesky’s privacy and rate‑limit policies.

Key capabilities include:

• Authentication: Automatic login using environment variables, with a fallback tool for alternate credentials or manual re‑authentication.
• Content creation: supports plain text and optional image uploads encoded in Base64, enabling media‑rich posts without manual multipart handling.
• Post management: Retrieve single or multiple posts, delete content, and manage reposts with dedicated tools.
• Interaction primitives: Like, unlike, repost, and unrepost actions are exposed as discrete tools, letting an agent react to trending content or curate its own feed.
• Profile and timeline access: and tools provide user data and a stream of recent posts, which can be fed into downstream processing or analytics pipelines.
• Prompt‑based formatting: The prompt turns raw timeline JSON into a human‑readable summary, useful for generating conversational responses or reports.

In real‑world scenarios, a marketing bot could automatically share campaign updates at optimal times, while a research assistant might monitor specific hashtags and summarize emerging discussions. Because the server is MCP‑compatible, any Claude client (Desktop or Web) can integrate these tools with minimal configuration—just point the MCP URL and provide credentials. This tight integration streamlines workflows, reduces boilerplate code, and ensures that AI agents can publish and interact on Bluesky as naturally as they converse.