Package Exports
- ai-debug-local-mcp
- ai-debug-local-mcp/dist/server.js
This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (ai-debug-local-mcp) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.
Readme
AI-Debug MCP - 318 Tools for Comprehensive Debugging
One Server. All Tools. Zero Complexity.
Quick Start
npm install -g ai-debug-local-mcp
Then configure your AI assistant (Claude, Cursor, Windsurf) to use the MCP server.
What You Get: 318 Real Tools
Core Debugging (40+ tools)
inject_debugging
- Start debugging any web pagetake_screenshot
- Capture page screenshotsget_console_logs
- Retrieve browser console outputmonitor_realtime
- Real-time event monitoringsimulate_user_action
- Click, type, scroll programmatically- And many more...
TDD & Testing (44 tools)
enable_tdd_mode
- Test-driven development with runtime bridgingrun_tests_with_coverage
- Execute tests with coverage analysistest_impact_analysis
- Understand change impactsregression_detection
- Catch regressions before deploymentfull_stack_tdd_workflow
- Backend changes with frontend validation
Framework-Specific Tools
Phoenix/Elixir (77 tools!)
- LiveView debugging and state inspection
- GenServer monitoring and tracing
- Ecto query analysis
- PubSub monitoring
- Process inspection
- And much more...
Flutter (15+ tools)
- Quantum debugging
- Widget tree analysis
- Accessibility testing
- Performance monitoring
Next.js (21 tools)
- SSR/SSG analysis
- Cache monitoring
- Bundle analysis
- Performance optimization
AI & Analysis (36 tools)
- AI-powered debugging assistance
- Token usage analysis
- Intelligent error analysis
- Feedback collection
Performance (20+ tools)
- Performance profiling
- Memory analysis
- Network monitoring
- Bottleneck detection
Quality & Standards (30+ tools)
- Accessibility auditing
- Security scanning
- Code quality gates
- Visual regression testing
And Many More Categories...
- GraphQL debugging (10 tools)
- Python backend (6 tools)
- Meta frameworks (10 tools)
- Circuit breaker management (6 tools)
- Visual testing (15+ tools)
Simple Architecture
One server. One command. All tools work.
Your AI Assistant <-> MCP Protocol <-> AI-Debug Server <-> Your Web App
Configuration
For Claude Desktop
Add to your Claude configuration:
{
"mcpServers": {
"ai-debug": {
"command": "ai-debug-mcp"
}
}
}
For Other AI Assistants
Most AI coding assistants support MCP. Check their documentation for configuration.
Examples
// Start debugging
await inject_debugging({ url: "http://localhost:3000" });
// Take a screenshot
await take_screenshot({ sessionId: "debug-123", fullPage: true });
// Run tests with coverage
await run_tests_with_coverage({ sessionId: "tdd-456", testPattern: "*.test.js" });
// Analyze Phoenix LiveView
await debug_liveview_state({ sessionId: "phoenix-789" });
Support
- GitHub: https://github.com/ai-debug/ai-debug-local-mcp
- Issues: https://github.com/ai-debug/ai-debug-local-mcp/issues
License
MIT