# Claude Skills MCP Backend
Heavy backend server for Claude Skills MCP system with vector search capabilities.
## Overview
This is the backend component of the Claude Skills MCP system. It provides:
- Vector-based semantic search using sentence-transformers
- Skill indexing and retrieval
- MCP protocol via Streamable HTTP transport
- Background skill loading from GitHub and local sources
**Note**: This package is typically auto-installed by the frontend (`claude-skills-mcp`). You only need to install it manually for:
- Remote deployment (hosting your own backend)
- Development and testing
- Standalone usage without the frontend proxy
## Installation
```bash
# Via uv (recommended)
uv tool install claude-skills-mcp-backend
# Via uvx (one-time use)
uvx claude-skills-mcp-backend
# Via pip
pip install claude-skills-mcp-backend
```
## Usage
### Run Standalone Server
```bash
# Default (localhost:8765)
claude-skills-mcp-backend
# Custom port
claude-skills-mcp-backend --port 8080
# For remote access
claude-skills-mcp-backend --host 0.0.0.0 --port 8080
# With custom configuration
claude-skills-mcp-backend --config my-config.json
# Verbose logging
claude-skills-mcp-backend --verbose
```
### Configuration
```bash
# Print example configuration
claude-skills-mcp-backend --example-config > config.json
# Edit config.json to customize skill sources, embedding model, etc.
# Run with custom config
claude-skills-mcp-backend --config config.json
```
## Endpoints
When running, the backend exposes:
- **Streamable HTTP MCP**: `http://localhost:8765/mcp`
- **Health Check**: `http://localhost:8765/health`
## Docker Deployment
### Build Image
```bash
docker build -t claude-skills-mcp-backend .
```
### Run Container
```bash
# For local access
docker run -p 8765:8765 claude-skills-mcp-backend
# For remote access
docker run -p 8080:8765 \
-e HOST=0.0.0.0 \
claude-skills-mcp-backend --host 0.0.0.0 --port 8765
```
## Dependencies
This package includes heavy dependencies (~250 MB):
- PyTorch (CPU-only on Linux): ~150-200 MB
- sentence-transformers: ~50 MB
- numpy, httpx, fastapi, uvicorn: ~30 MB
**First download may take 60-180 seconds** depending on your internet connection.
## Performance
- **Startup time**: 2-5 seconds (with cached dependencies)
- **First search**: +2-5 seconds (embedding model download, one-time)
- **Query time**: <1 second after models loaded
- **Memory usage**: ~500 MB
## Development
```bash
# Clone the monorepo
git clone https://github.com/K-Dense-AI/claude-skills-mcp.git
cd claude-skills-mcp/packages/backend
# Install in development mode
uv pip install -e ".[test]"
# Run tests
uv run pytest tests/
```
## Related Packages
- **claude-skills-mcp** (Frontend): Lightweight proxy that auto-installs this backend
- **Main Repository**: https://github.com/K-Dense-AI/claude-skills-mcp
## License
Apache License 2.0
Copyright 2025 K-Dense AI (https://k-dense.ai)
Raw data
{
"_id": null,
"home_page": null,
"name": "claude-skills-mcp-backend",
"maintainer": null,
"docs_url": null,
"requires_python": "<3.13,>=3.12",
"maintainer_email": "K-Dense AI <orion.li@k-dense.ai>",
"keywords": "agent, ai, backend, claude, mcp, scientific, skills, vector-search",
"author": null,
"author_email": "K-Dense AI <orion.li@k-dense.ai>",
"download_url": "https://files.pythonhosted.org/packages/5e/91/27acabbf6c890627caf28164ac230f9de57c8a73f8e99dacab3c90779e60/claude_skills_mcp_backend-1.0.6.tar.gz",
"platform": null,
"description": "# Claude Skills MCP Backend\n\nHeavy backend server for Claude Skills MCP system with vector search capabilities.\n\n## Overview\n\nThis is the backend component of the Claude Skills MCP system. It provides:\n- Vector-based semantic search using sentence-transformers\n- Skill indexing and retrieval\n- MCP protocol via Streamable HTTP transport\n- Background skill loading from GitHub and local sources\n\n**Note**: This package is typically auto-installed by the frontend (`claude-skills-mcp`). You only need to install it manually for:\n- Remote deployment (hosting your own backend)\n- Development and testing\n- Standalone usage without the frontend proxy\n\n## Installation\n\n```bash\n# Via uv (recommended)\nuv tool install claude-skills-mcp-backend\n\n# Via uvx (one-time use)\nuvx claude-skills-mcp-backend\n\n# Via pip\npip install claude-skills-mcp-backend\n```\n\n## Usage\n\n### Run Standalone Server\n\n```bash\n# Default (localhost:8765)\nclaude-skills-mcp-backend\n\n# Custom port\nclaude-skills-mcp-backend --port 8080\n\n# For remote access\nclaude-skills-mcp-backend --host 0.0.0.0 --port 8080\n\n# With custom configuration\nclaude-skills-mcp-backend --config my-config.json\n\n# Verbose logging\nclaude-skills-mcp-backend --verbose\n```\n\n### Configuration\n\n```bash\n# Print example configuration\nclaude-skills-mcp-backend --example-config > config.json\n\n# Edit config.json to customize skill sources, embedding model, etc.\n\n# Run with custom config\nclaude-skills-mcp-backend --config config.json\n```\n\n## Endpoints\n\nWhen running, the backend exposes:\n\n- **Streamable HTTP MCP**: `http://localhost:8765/mcp`\n- **Health Check**: `http://localhost:8765/health`\n\n## Docker Deployment\n\n### Build Image\n\n```bash\ndocker build -t claude-skills-mcp-backend .\n```\n\n### Run Container\n\n```bash\n# For local access\ndocker run -p 8765:8765 claude-skills-mcp-backend\n\n# For remote access\ndocker run -p 8080:8765 \\\n -e HOST=0.0.0.0 \\\n claude-skills-mcp-backend --host 0.0.0.0 --port 8765\n```\n\n## Dependencies\n\nThis package includes heavy dependencies (~250 MB):\n- PyTorch (CPU-only on Linux): ~150-200 MB\n- sentence-transformers: ~50 MB\n- numpy, httpx, fastapi, uvicorn: ~30 MB\n\n**First download may take 60-180 seconds** depending on your internet connection.\n\n## Performance\n\n- **Startup time**: 2-5 seconds (with cached dependencies)\n- **First search**: +2-5 seconds (embedding model download, one-time)\n- **Query time**: <1 second after models loaded\n- **Memory usage**: ~500 MB\n\n## Development\n\n```bash\n# Clone the monorepo\ngit clone https://github.com/K-Dense-AI/claude-skills-mcp.git\ncd claude-skills-mcp/packages/backend\n\n# Install in development mode\nuv pip install -e \".[test]\"\n\n# Run tests\nuv run pytest tests/\n```\n\n## Related Packages\n\n- **claude-skills-mcp** (Frontend): Lightweight proxy that auto-installs this backend\n- **Main Repository**: https://github.com/K-Dense-AI/claude-skills-mcp\n\n## License\n\nApache License 2.0\n\nCopyright 2025 K-Dense AI (https://k-dense.ai)\n\n",
"bugtrack_url": null,
"license": "Apache-2.0",
"summary": "Heavy backend for Claude Skills MCP server with vector search (Streamable HTTP)",
"version": "1.0.6",
"project_urls": {
"Bug Tracker": "https://github.com/K-Dense-AI/claude-skills-mcp/issues",
"Documentation": "https://github.com/K-Dense-AI/claude-skills-mcp/blob/main/packages/backend/README.md",
"Homepage": "https://github.com/K-Dense-AI/claude-skills-mcp",
"Repository": "https://github.com/K-Dense-AI/claude-skills-mcp"
},
"split_keywords": [
"agent",
" ai",
" backend",
" claude",
" mcp",
" scientific",
" skills",
" vector-search"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "3ef81fbac56a2330bc35bc5b288f67b1ae8586565a9d5e78dc9cdfb1abbbfcde",
"md5": "331fecbc0417a4e51abd7a0512945ef9",
"sha256": "7c522651c5bbffbdecd1200ad08fb009cf45cfc5f23281a99cafd78ca1362afa"
},
"downloads": -1,
"filename": "claude_skills_mcp_backend-1.0.6-py3-none-any.whl",
"has_sig": false,
"md5_digest": "331fecbc0417a4e51abd7a0512945ef9",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": "<3.13,>=3.12",
"size": 31423,
"upload_time": "2025-10-24T01:19:49",
"upload_time_iso_8601": "2025-10-24T01:19:49.095438Z",
"url": "https://files.pythonhosted.org/packages/3e/f8/1fbac56a2330bc35bc5b288f67b1ae8586565a9d5e78dc9cdfb1abbbfcde/claude_skills_mcp_backend-1.0.6-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "5e9127acabbf6c890627caf28164ac230f9de57c8a73f8e99dacab3c90779e60",
"md5": "b3d4825b5a0aa2097b563e7c16e896df",
"sha256": "6954ea824a15182c485265e32709ebccce73516d0f3353d5fc983cc8952cc0f4"
},
"downloads": -1,
"filename": "claude_skills_mcp_backend-1.0.6.tar.gz",
"has_sig": false,
"md5_digest": "b3d4825b5a0aa2097b563e7c16e896df",
"packagetype": "sdist",
"python_version": "source",
"requires_python": "<3.13,>=3.12",
"size": 78345,
"upload_time": "2025-10-24T01:19:50",
"upload_time_iso_8601": "2025-10-24T01:19:50.386929Z",
"url": "https://files.pythonhosted.org/packages/5e/91/27acabbf6c890627caf28164ac230f9de57c8a73f8e99dacab3c90779e60/claude_skills_mcp_backend-1.0.6.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-10-24 01:19:50",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "K-Dense-AI",
"github_project": "claude-skills-mcp",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "claude-skills-mcp-backend"
}
JSON
