{"_id":"@dylannwatt/mcp-coding-agents","name":"@dylannwatt/mcp-coding-agents","dist-tags":{"latest":"1.0.0"},"versions":{"1.0.0":{"name":"@dylannwatt/mcp-coding-agents","version":"1.0.0","main":"dist/index.js","bin":{"mcp-coding-agents":"dist/cli-unified.js"},"scripts":{"build":"tsc","start":"node dist/index.js","dev":"nodemon --exec ts-node src/index.ts","dev:watch":"tsc --watch","mcp":"node dist/cli-unified.js","mcp:dev":"ts-node src/cli-unified.ts","mcp:stdio":"node dist/cli-unified.js stdio","mcp:stdio:dev":"ts-node src/cli-unified.ts stdio","mcp:http":"node dist/cli-unified.js http","mcp:http:dev":"ts-node src/cli-unified.ts http","mcp:legacy":"node dist/mcp-server-http.js","test":"jest","test:watch":"jest --watch","test:coverage":"jest --coverage","test:integration":"jest tests/integration"},"author":{"name":"kardio"},"license":"MIT","keywords":["mcp","model-context-protocol","claude","ai","server"],"description":"MCP server with multiple AI coding agents for enhanced development workflows","repository":{"type":"git","url":"git+https://github.com/kadreio/mcp-coding-agents.git"},"engines":{"node":">=18.0.0"},"dependencies":{"@anthropic-ai/claude-code":"^1.0.58","@modelcontextprotocol/sdk":"^1.15.1","commander":"^14.0.0","dotenv":"^17.2.0","express":"^5.1.0"},"devDependencies":{"@types/express":"^5.0.3","@types/jest":"^30.0.0","@types/node":"^24.0.14","@types/supertest":"^6.0.3","jest":"^29.7.0","nodemon":"^3.1.10","supertest":"^7.1.3","ts-jest":"^29.4.0","ts-node":"^10.9.2","typescript":"^5.8.3"},"_id":"@dylannwatt/mcp-coding-agents@1.0.0","gitHead":"1d5b9b591702e247622721ba5d73dbf726f6b3b2","bugs":{"url":"https://github.com/kadreio/mcp-coding-agents/issues"},"homepage":"https://github.com/kadreio/mcp-coding-agents#readme","_nodeVersion":"23.11.1","_npmVersion":"10.9.2","dist":{"integrity":"sha512-GxCZ6+jglMsmALJA2nCK8a7r+5cJVYYij88/LapcA62r5QHBR7ceyPAhM57OHrUVyP9Rx6VNMIQdi0LhmGcA7w==","shasum":"23a23d2e3a97879cdaa542ba1f1cbef0314afd7a","tarball":"https://registry.npmjs.org/@dylannwatt/mcp-coding-agents/-/mcp-coding-agents-1.0.0.tgz","fileCount":16,"unpackedSize":100469,"signatures":[{"keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U","sig":"MEUCIBFlJwQIiL+oCHTSfbUBxDs6P7xubX5Jv7lEzNSAWocKAiEA23/nDgGYtdIJhV8Y2yyG53XEHzQHtTTzP6o6oz3RXRs="}]},"_npmUser":{"name":"dylannwatt","email":"dylannwatt@gmail.com"},"directories":{},"maintainers":[{"name":"dylannwatt","email":"dylannwatt@gmail.com"}],"_npmOperationalInternal":{"host":"s3://npm-registry-packages-npm-production","tmp":"tmp/mcp-coding-agents_1.0.0_1754255656271_0.381288642765925"},"_hasShrinkwrap":false}},"time":{"created":"2025-08-03T21:14:16.192Z","1.0.0":"2025-08-03T21:14:16.449Z","modified":"2025-08-03T21:14:16.713Z"},"maintainers":[{"name":"dylannwatt","email":"dylannwatt@gmail.com"}],"description":"MCP server with multiple AI coding agents for enhanced development workflows","homepage":"https://github.com/kadreio/mcp-coding-agents#readme","keywords":["mcp","model-context-protocol","claude","ai","server"],"repository":{"type":"git","url":"git+https://github.com/kadreio/mcp-coding-agents.git"},"author":{"name":"kardio"},"bugs":{"url":"https://github.com/kadreio/mcp-coding-agents/issues"},"license":"MIT","readme":"# @kadreio/mcp-claude-code\n\nA transport-agnostic MCP (Model Context Protocol) server with Claude Code, Gemini, and Codex agent integrations. Supports both STDIO and HTTP transports for flexible AI model context management.\n\n## Features\n\n- 🚀 **Transport-agnostic architecture** - Clean separation between business logic and transport layers\n- 🔄 **Multiple transport modes** - STDIO for direct process communication, HTTP with SSE for network access\n- 🤖 **Multiple AI agents** - Claude Code, Gemini, and Codex integrations\n- 📦 **Built-in tools** - BMI calculator, timestamp generator, command execution, and more\n- 🔧 **Extensible design** - Easy to add new tools, prompts, and transports\n- 💬 **Real-time streaming** - Support for notifications and SSE in HTTP mode\n\n## Installation\n\n```bash\nnpm install -g @kadreio/mcp-claude-code\n```\n\n## Quick Start\n\n### Using npx (no installation required)\n\n```bash\n# Default mode (HTTP on port 3050)\nnpx @kadreio/mcp-claude-code\n\n# STDIO mode for direct process communication\nnpx @kadreio/mcp-claude-code stdio\n\n# HTTP mode with custom port\nnpx @kadreio/mcp-claude-code http --port 3051\n\n# With specific host binding\nnpx @kadreio/mcp-claude-code http --host 0.0.0.0\n```\n\n### After global installation\n\n```bash\n# Default HTTP mode\nmcp-claude-code\n\n# STDIO mode\nmcp-claude-code stdio\n\n# HTTP mode with options\nmcp-claude-code http --port 3051 --no-cors\n```\n\n## Transport Modes\n\n### STDIO Mode\nBest for editor integrations and direct process communication:\n```bash\nmcp-claude-code stdio\n```\n- Communicates via stdin/stdout using JSON-RPC\n- No network overhead\n- Ideal for local development tools\n\n### HTTP Mode\nBest for network access and web integrations:\n```bash\nmcp-claude-code http --port 3050\n```\n- RESTful API with SSE support for streaming\n- Session management for stateful interactions\n- CORS support (configurable)\n- Health check endpoint at `/health`\n\n## Available Tools\n\n### Core Tools\n- `calculate_bmi` - Calculate Body Mass Index\n- `get_timestamp` - Get current timestamp\n- `execute_command` - Execute shell commands with timeout support\n- `stream_sse_timestamps` - Generate streaming timestamps (HTTP mode)\n\n### AI Agent Tools\n- `claude_code_query` - Query Claude Code with streaming responses\n- `gemini_query` - Execute Gemini CLI commands\n- `codex_query` - Execute Codex with JSONL streaming\n\n## Environment Variables\n\n```bash\n# Core Configuration\nMCP_PORT=3050                    # HTTP server port (default: 3050)\n\n# Claude Code Configuration\nANTHROPIC_API_KEY=sk-ant-...     # Required for Claude Code\nCLAUDE_CODE_ENABLE=true          # Enable/disable Claude Code tool\nCLAUDE_CODE_DEFAULT_CWD=/path    # Default working directory\nCLAUDE_CODE_DEFAULT_MODEL=...    # Default Claude model\nCLAUDE_CODE_MAX_MESSAGES=100     # Max messages to return\n\n# Agent Configuration\nGEMINI_API_KEY=...               # For Gemini agent\nOPENAI_API_KEY=...               # For Codex agent\n```\n\n## Development\n\n```bash\n# Clone the repository\ngit clone https://github.com/kadreio/mcp-claude-code.git\ncd mcp-claude-code\n\n# Install dependencies\nnpm install\n\n# Build the project\nnpm run build\n\n# Development mode with hot reload\nnpm run mcp:dev          # Default HTTP mode\nnpm run mcp:stdio:dev    # STDIO mode\nnpm run mcp:http:dev     # Explicit HTTP mode\n\n# Run tests\nnpm test\nnpm run test:coverage\n```\n\n## Architecture\n\nThe server uses a clean, transport-agnostic architecture:\n\n```\nCoreMCPServer (Business Logic)\n       ↓\nTransport Interface\n    ↙     ↘\nSTDIO    HTTP\nTransport Transport\n```\n\n- **CoreMCPServer**: Contains all business logic, tool handlers, and agent integrations\n- **Transport Layer**: Thin adapters for different communication protocols\n- **Unified CLI**: Single entry point with transport selection\n\n## API Endpoints (HTTP Mode)\n\n- `POST /mcp` - Main MCP endpoint for JSON-RPC requests\n- `GET /mcp` - SSE endpoint for streaming notifications\n- `GET /health` - Health check endpoint\n- `GET /` - Server information and capabilities\n\n## MCP Client Configuration\n\n### For STDIO mode\n```json\n{\n  \"mcpServers\": {\n    \"claude-code\": {\n      \"command\": \"npx\",\n      \"args\": [\"@kadreio/mcp-claude-code\", \"stdio\"],\n      \"env\": {\n        \"ANTHROPIC_API_KEY\": \"your-api-key\"\n      }\n    }\n  }\n}\n```\n\n### For HTTP mode\n```json\n{\n  \"mcpServers\": {\n    \"claude-code\": {\n      \"url\": \"http://localhost:3050/mcp\",\n      \"transport\": \"http\"\n    }\n  }\n}\n```\n\n## Contributing\n\nSee [CLAUDE.md](./CLAUDE.md) for development guidelines and architecture details.\n\n## License\n\nMIT","readmeFilename":"README.md","_rev":"1-5752e702fd904f4aa051463b2989e8ba"}