Spaces:
Runtime error
Runtime error
|
Download mcp-server-setup.md from BTankut/Context-Store-LLM: direct link, hf CLI and curl.
- Browser
- Download file 5.53 kB
-
https://huggingface.co/spaces/BTankut/Context-Store-LLM/resolve/main/mcp-server-setup.md
- Command line
-
hf download hf://spaces/BTankut/Context-Store-LLM/mcp-server-setup.md
-
curl -L -o mcp-server-setup.md https://huggingface.co/spaces/BTankut/Context-Store-LLM/resolve/main/mcp-server-setup.md
5.53 kB
| # Context Store MCP Server Setup | |
| ## Cloud MCP Server Configuration | |
| Once Context Store is deployed to HuggingFace, you can access it from any device using the MCP protocol. | |
| ### Add to Claude Code | |
| ```bash | |
| # Add Context Store MCP server | |
| claude mcp add context-store https://your-username-context-store.hf.space | |
| # Verify connection | |
| claude mcp list | |
| ``` | |
| ### MCP Configuration File | |
| Create or update `~/.claude/mcp.json`: | |
| ```json | |
| { | |
| "servers": { | |
| "context-store": { | |
| "command": "node", | |
| "args": ["-e", " | |
| const https = require('https'); | |
| const readline = require('readline'); | |
| const baseUrl = 'https://your-username-context-store.hf.space'; | |
| const rl = readline.createInterface({ | |
| input: process.stdin, | |
| output: process.stdout | |
| }); | |
| rl.on('line', async (input) => { | |
| try { | |
| const request = JSON.parse(input); | |
| if (request.method === 'tools/list') { | |
| console.log(JSON.stringify({ | |
| jsonrpc: '2.0', | |
| id: request.id, | |
| result: { | |
| tools: [ | |
| { | |
| name: 'add_memory', | |
| description: 'Add content to Context Store memory', | |
| inputSchema: { | |
| type: 'object', | |
| properties: { | |
| content: { type: 'string', description: 'Content to store' }, | |
| source: { type: 'string', description: 'Source identifier' } | |
| }, | |
| required: ['content'] | |
| } | |
| }, | |
| { | |
| name: 'search_memory', | |
| description: 'Search Context Store memory', | |
| inputSchema: { | |
| type: 'object', | |
| properties: { | |
| query: { type: 'string', description: 'Search query' }, | |
| limit: { type: 'number', description: 'Max results' } | |
| }, | |
| required: ['query'] | |
| } | |
| } | |
| ] | |
| } | |
| })); | |
| } | |
| else if (request.method === 'tools/call') { | |
| const { name, arguments: args } = request.params; | |
| if (name === 'add_memory') { | |
| const response = await fetch(`${baseUrl}/api/v1/add`, { | |
| method: 'POST', | |
| headers: { 'Content-Type': 'application/json' }, | |
| body: JSON.stringify({ | |
| episodeBody: args.content, | |
| source: args.source || 'mcp' | |
| }) | |
| }); | |
| const data = await response.json(); | |
| console.log(JSON.stringify({ | |
| jsonrpc: '2.0', | |
| id: request.id, | |
| result: { content: [{ type: 'text', text: JSON.stringify(data) }] } | |
| })); | |
| } | |
| else if (name === 'search_memory') { | |
| const response = await fetch(`${baseUrl}/api/v1/search?q=${encodeURIComponent(args.query)}&limit=${args.limit || 10}`); | |
| const data = await response.json(); | |
| console.log(JSON.stringify({ | |
| jsonrpc: '2.0', | |
| id: request.id, | |
| result: { content: [{ type: 'text', text: JSON.stringify(data) }] } | |
| })); | |
| } | |
| } | |
| } catch (error) { | |
| console.log(JSON.stringify({ | |
| jsonrpc: '2.0', | |
| id: request.id, | |
| error: { code: -1, message: error.message } | |
| })); | |
| } | |
| }); | |
| "] | |
| } | |
| } | |
| } | |
| ``` | |
| ### Alternative: Use Claude Code Built-in MCP | |
| If using Claude Code's MCP capabilities: | |
| ```bash | |
| # Direct HTTP MCP connection | |
| claude config set mcp.context-store.url "https://your-username-context-store.hf.space" | |
| claude config set mcp.context-store.type "http" | |
| # Test connection | |
| claude mcp test context-store | |
| ``` | |
| ## Usage Examples | |
| ### Add Memory | |
| ```bash | |
| claude -m "Add this conversation to my Context Store" --mcp context-store | |
| ``` | |
| ### Search Memory | |
| ```bash | |
| claude -m "Search my Context Store for React patterns" --mcp context-store | |
| ``` | |
| ### Automatic Memory | |
| ```bash | |
| # Configure auto-save to Context Store | |
| claude config set memory.provider "context-store" | |
| claude config set memory.auto-save true | |
| ``` | |
| ## Device Setup | |
| ### Windows | |
| ```cmd | |
| # Install Claude Code | |
| npm install -g @anthropic/claude-code | |
| # Add MCP server | |
| claude mcp add context-store https://your-space.hf.space | |
| ``` | |
| ### macOS/Linux | |
| ```bash | |
| # Install Claude Code | |
| curl -fsSL https://claude.ai/install.sh | sh | |
| # Add MCP server | |
| claude mcp add context-store https://your-space.hf.space | |
| ``` | |
| ### Mobile (Termux/iSH) | |
| ```bash | |
| # Install Node.js and Claude Code | |
| pkg install nodejs-lts | |
| npm install -g @anthropic/claude-code | |
| # Configure MCP | |
| claude mcp add context-store https://your-space.hf.space | |
| ``` | |
| ## Benefits | |
| ✅ **Universal Access**: Same memory on all devices | |
| ✅ **No Local Setup**: No Docker, databases, or services needed | |
| ✅ **Always Available**: HuggingFace handles uptime and scaling | |
| ✅ **Free Tier**: No hosting costs for development usage | |
| ✅ **MCP Protocol**: Standard AI tool integration | |
| ✅ **Real-time Sync**: Changes available immediately everywhere |