aider-mcp-client
A Python tool to query Context7 MCP for library docs. Integrates with Aider for AI coding. Install via pip install mcp_query and fetch docs easily.
GitHubスター
2
ユーザー評価
未評価
お気に入り
0
閲覧数
25
フォーク
0
イシュー
0
Aider MCP Client: Supercharge Your Coding with AI-Powered Documentation
A Python client that seamlessly integrates MCP (Model Control Protocol) with Aider, creating a powerful AI-assisted development workflow. Fetch precise documentation snippets directly into your coding sessions while efficiently managing token usage.
Why Use MCP with Aider?
- Context-Aware Coding: Get relevant documentation snippets injected directly into your AI coding sessions
- Token Efficiency: Aider's smart token management ensures you get maximum context without wasting tokens
- Reduced Hallucinations: Ground your AI coding in actual documentation rather than made-up examples
- Faster Development: Eliminate context-switching between docs and code
- Precision Answers: Get documentation tailored to your exact coding context
Key Features
- Simple configuration via JSON
- Command-line interface
- Aider-compatible JSON output
- Integration with Context7 MCP servers
Smart Token Management with Aider
Aider was chosen as the foundation for this client because of its excellent token management capabilities:
- Automatic Token Optimization: Aider intelligently manages context window size
- Prioritized Context: Keeps the most relevant documentation in the prompt
- Documentation Chunking: Breaks large docs into manageable pieces
- Session Awareness: Maintains context across multiple queries
This ensures you get the most value from your AI coding sessions without hitting token limits.
Installation
From PyPI:
pip install aider-mcp-client
From GitHub:
pip install git+https://github.com/alvinveroy/aider-mcp-client.git
For development:
git clone https://github.com/alvinveroy/aider-mcp-client.git
cd aider-mcp-client
pip install -e .
Usage
After installation, you can use the command-line interface:
aider_mcp_client <command> [options] [args...]
Or as a module:
python -m aider_mcp_client <command> [options] [args...]
Command-line Options
usage: aider_mcp_client [-h] [-v] [--debug] [--verbose] [--quiet] [--server SERVER] [--json]
{fetch,resolve,list} ...
Aider MCP client for fetching library documentation.
positional arguments:
{fetch,resolve,list} Available commands
fetch Fetch documentation for a library
resolve Resolve a library name to a Context7-compatible ID
list List supported libraries
options:
-h, --help show this help message and exit
-v, --version Show version information
--debug Enable debug logging
--verbose Show detailed logs in console
--quiet Suppress informational output
--server SERVER MCP server to use (default: context7)
--json Force JSON output format
Example Workflows
Basic Documentation Fetch
# Get React hooks documentation (automatically optimized for token usage)
aider_mcp_client fetch react --topic "hooks"
# Get Next.js routing docs with 10k token budget
aider_mcp_client fetch vercel/nextjs --topic "routing" --tokens 10000
Integration with Aider
# Start an Aider session with React context
aider --context $(aider_mcp_client fetch react --json)
Advanced Usage
# Resolve library name to ID (useful for scripting)
aider_mcp_client resolve next.js
# Get docs in JSON format for processing
aider_mcp_client fetch react --topic "state" --json
# Debug connection issues
aider_mcp_client fetch react --verbose --debug
Configuration
The client uses a configuration file located at ~/.aider-mcp-client/config.json
. If this file doesn't exist, default settings are used.
Default configuration:
{
"mcpServers": {
"context7": {
"command": "npx",
"args": ["-y", "@upstash/context7-mcp@latest"],
"enabled": true,
"timeout": 30
}
}
}
You can create a custom configuration file:
mkdir -p ~/.aider-mcp-client
echo '{
"mcpServers": {
"context7": {
"command": "npx",
"args": ["-y", "@upstash/context7-mcp@latest"],
"enabled": true,
"timeout": 30
}
}
}' > ~/.aider-mcp-client/config.json
Available Commands
version
: Display version informationaider_mcp_client # or aider_mcp_client -v # or aider_mcp_client --version
fetch
: Retrieve documentation for a specific libraryaider_mcp_client fetch <library_id> [--topic "topic"] [--tokens 5000]
or
python -m aider_mcp_client.client fetch <library_id> [--topic "topic"] [--tokens 5000]
Aider Integration
The client outputs JSON in Aider-compatible format:
{
"content": "...",
"library": "library_name",
"snippets": [...],
"totalTokens": 5000,
"lastUpdated": "timestamp"
}
Development
Running Tests
# Install pytest first
pip install pytest
# Then run the tests
python -m pytest tests/
You can also use unittest if you prefer:
python -m unittest discover tests
Code Structure
aider_mcp_client/client.py
: Main client implementation with CLI interfaceaider_mcp_client/config.json
: Default configuration templatetests/
: Unit tests
Contributing
- Fork the repository
- Create a feature branch
- Submit a pull request
License
MIT - See LICENSE for details.