BrianKnows MCP Server
# BrianKnows MCP Server
[](https://smithery.ai/server/@antoncoding/mcp-brianknows)
A Model Context Protocol (MCP) server that connects Claude to BrianKnows' blockchain knowledge base.
<a href="https://glama.ai/mcp/servers/idfph0fstx">
<img width="380" height="200" src="https://glama.ai/mcp/servers/idfph0fstx/badge" alt="BrianKnows Server MCP server" />
</a>
## What is MCP? 🤔
The Model Context Protocol (MCP) lets AI assistants like Claude Desktop connect to external tools and data sources in a secure way while keeping users in control.
## What does this server do? 🚀
The BrianKnows MCP server provides three main tools:
1. **Ping Tool**: Check if the BrianKnows API server is responsive
2. **Search Tool**: Query BrianKnows' knowledge engine for blockchain and DeFi information
3. **Agent Tool**: Chat with the BrianKnows agent about DeFi protocols
Supported knowledge bases include:
- public-knowledge-box (default)
- circle_kb, lido_kb, Polygon_kb, taiko_kb
- near_kb, clave_kb, starknet_kb, consensys_kb
The server maintains a cache of your 5 most recent searches for quick reference.
## Prerequisites 📋
- [Node.js](https://nodejs.org/) (v18 or higher)
- [Claude Desktop](https://claude.ai/download)
- A [BrianKnows API key](https://docs.brianknows.org/)
## Configuration ⚙️
Add this to your Claude Desktop configuration file (accessible via Developer Settings):
```json
{
"mcpServers": {
"brianknows": {
"command": "npx",
"args": ["mcp-brianknows"],
"env": {
"BRIAN_API_KEY": "your-api-key-here"
}
}
}
}
```
Replace `your-api-key-here` with your actual BrianKnows API key.
## Example Usage 🎯
```
Can you check if the BrianKnows API is online?
Use BrianKnows to search for information about Ethereum's Layer 2 solutions.
Ask the BrianKnows agent to explain how Uniswap V3 works.
```
## Features ✨
* **Multiple Knowledge Bases**: Access specialized knowledge for different blockchain protocols
* **Cached Searches**: Quick access to your 5 most recent searches
* **Error Handling**: User-friendly error messages
* **Type Safety**: Full TypeScript implementation
## Acknowledgments 🙏
* [BrianKnows](https://brianknows.org) for their blockchain knowledge API
* [Model Context Protocol](https://modelcontextprotocol.io)
* [Anthropic](https://anthropic.com) for Claude DesktopTDQS
Scored across 3 tools
Each tool has a clearly distinct purpose with no overlap: 'agent' is for chatting with an AI agent, 'ping' is for health checking the API server, and 'search' is for querying a knowledge engine. The descriptions make it easy to differentiate these functions, eliminating any risk of misselection.
The naming is mixed in style: 'agent' and 'ping' are simple verbs, while 'search' is a verb that could fit a pattern, but there is no consistent verb_noun structure or uniform convention. However, the names are readable and straightforward, avoiding chaotic or confusing formats.
With only 3 tools, the count feels thin for a server named 'BrianKnows MCP Server', which suggests a broader knowledge or interaction domain. While the tools cover basic operations (chat, health check, search), it may lack depth for more complex workflows, making it borderline in scope.
The tool surface has significant gaps for a knowledge-oriented server: there is no way to create, update, or manage knowledge entries, and the 'agent' tool lacks clarity on its capabilities beyond chatting. This incompleteness could lead to agent failures when trying to perform comprehensive tasks in the domain.