Today, I'm excited to dive deeper into our latest announcement about Needle's MCP (Model Context Protocol) capabilities that we've just rolled out. If you haven't seen the video walkthrough yet, I encourage you to check it out for a complete demonstration.
What We've Released
We've launched two complementary MCP features that dramatically expand how you can use Needle in your AI workflow:
Needle as an MCP Server: Expose Needle's powerful semantic search capabilities to any MCP client
Needle as an MCP Client: Connect Needle to other tools' MCP servers
This dual approach creates a flexible ecosystem that puts Needle at the center of your AI tool stack.
Needle MCP Server: Semantic Search Everywhere
Our remote MCP server is now available at https://mcp.needle-ai.com/sse, accessible via either SSE transport or streamable HTTP https://mcp.needle-ai.com/mcp. This means you can enable Needle's semantic search/RAG capabilities in virtually any MCP-compatible client.
Key benefits:
Effortless setup: No need to install or maintain any local servers
Always current: Access the latest features without updates
Powerful RAG: Bring semantic search to your favorite AI tools
Once configured, you can instantly access your Needle collections and perform semantic searches through Claude's interface. The same works for any MCP-compatible client, including Cursor or your own custom applications.
Needle as an MCP Client: Bringing Tools Into Needle
The flip side is equally powerful… we've made Needle Web an MCP client that can connect to any MCP server. This means you can bring tools from other platforms directly into your Needle collections.
In the demo, I showed how to:
Create a new collection in Needle
Add a connector for Linear's MCP server
Authenticate with Linear via OAuth
Instantly gain access to Linear's issue tracking tools right inside Needle
This capability opens up exciting possibilities for creating specialized assistants that combine Needle's semantic search with domain-specific tools from other platforms.
The MCP Ecosystem Is Growing
The MCP standard is gaining momentum, with more tools providing MCP servers every week. Some notable examples include:
Linear (issue tracking)
Asana (project management)
Intercom (customer support)
PayPal (payments)
And many more
A comprehensive list is available in this GitHub repository.
What This Means For You
These new capabilities create a flexible foundation for AI tool integration:
Use Needle anywhere: Bring your collections and semantic search to other AI interfaces
Specialized assistants: Create purpose-built AI assistants within Needle that leverage external tools
Unified workflows: Seamlessly combine knowledge retrieval with specific actions in other platforms
Getting Started
To start using Needle as an MCP server:
Get your API key from the Settings page in Needle
Configure your MCP client with our server URL and your key
Start searching your collections from your preferred interface
To use external tools in Needle:
Create a collection or open an existing one
Add a new connector and select "MCP Server"
Enter the MCP server URL of the tool you want to integrate
Authenticate and start using the tool directly in your Needle chat
What's Next?
We're just scratching the surface of what's possible with MCP integration. Our team is already working on more connectors and expanded capabilities to make Needle an even more powerful hub for your AI-enhanced workflows.
I'd love to hear what you build with these new capabilities! Share your use cases in the comments or reach out to our team with any questions.
P.S. If you haven't tried Needle yet, sign up for free and experience the power of semantic search combined with seamless tool integration!
Share this post