Context7 MCP Server

This extension provides integration with Context7 MCP server for up-to-date, version-specific library documentation.

Features

• Resolve library identifiers for accurate documentation lookup
• Get current, version-specific documentation and code examples
• Reduce LLM hallucination by using real-time library docs instead of training data
• Support for thousands of libraries across multiple languages
• Optional API key for higher rate limits and private repository access

What is Context7?

Context7 is an MCP server that delivers current documentation directly into LLM prompts, addressing the problem that AI models rely on outdated training data. It prevents hallucinated APIs and enables accurate code generation.

Authentication

Context7 works without an API key (with standard rate limits). For higher limits and private repo access:

1. Get a free API key at context7.com/dashboard
2. Add to your sindri.yaml:

   secrets:
     - name: CONTEXT7_API_KEY
       source: env

3. Set the environment variable or add to .env file
4. Reinstall: extension-manager reinstall context7-mcp

Configuration

The Context7 MCP server is automatically added to your user-scope configuration at ~/.claude.json:

Without API key:

{
  "mcpServers": {
    "context7": {
      "type": "http",
      "url": "https://mcp.context7.com/mcp"
    }
  }
}

With API key (higher rate limits):

{
  "mcpServers": {
    "context7": {
      "type": "http",
      "url": "https://mcp.context7.com/mcp",
      "headers": {
        "CONTEXT7_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

Manual Installation

If automatic installation fails:

Without API key:

claude mcp add --transport http --scope user context7 https://mcp.context7.com/mcp

With API key:

claude mcp add --transport http --scope user --header "CONTEXT7_API_KEY: YOUR_KEY" context7 https://mcp.context7.com/mcp

Or using JSON:

claude mcp add-json --scope user context7 '{"type":"http","url":"https://mcp.context7.com/mcp"}'

Usage

Once installed, ask Claude:

• "What's the latest React documentation for hooks?"
• "Show me how to use pandas 2.0 DataFrame with examples"
• "Get FastAPI authentication documentation"
• "How do I use Next.js 14 app router?"
• "Find documentation for TypeScript 5.3 decorators"

Available Tools

The MCP server exposes these tools:

Tool	Description
resolve-library-id	Convert general library names to Context7 identifiers
get-library-docs	Retrieve version-specific docs with topic filtering

Tool Examples

resolve-library-id:

• Input: "react", "pandas", "fastapi"
• Output: Context7-compatible library identifier

get-library-docs:

• Parameters: library_id, topic (optional), page (1-10, optional)
• Returns: Up-to-date documentation and code examples

Rate Limits

Tier	Rate Limit	Cost	Use Case
No API key	Limited	Free	Personal projects, testing
Free API key	Higher	Free	Development, small teams
Paid tiers	Highest	Paid	Production, large teams

Validation

# Check Claude CLI is available
claude --version

# Verify Context7 MCP is configured
claude mcp list --scope user

# Get Context7 MCP details
claude mcp get context7

Troubleshooting

Server Not Responding

# Check status
claude mcp get context7

# Remove and reinstall
claude mcp remove --scope user context7
extension-manager reinstall context7-mcp

Add API Key After Installation

1. Get API key from context7.com/dashboard
2. Add to sindri.yaml secrets section
3. Set environment variable: export CONTEXT7_API_KEY=your_key
4. Reinstall: extension-manager reinstall context7-mcp

View Configuration

cat ~/.claude.json | jq '.mcpServers.context7'

Rate Limit Exceeded

If you see rate limit errors without an API key:

1. Get a free API key at context7.com/dashboard
2. Follow "Add API Key After Installation" steps above

Removal

extension-manager remove context7-mcp

This removes:

• Context7 MCP from Claude Code user configuration
• Extension directory at ~/extensions/context7-mcp

Links

• Context7 GitHub
• Context7 Dashboard
• MCP Protocol
• Claude Code MCP Docs

Related Extensions

• linear-mcp - Linear project management
• jira-mcp - Atlassian Jira/Confluence integration
• ai-toolkit - AI CLI tools suite
• ollama - Local LLM runtime