{"$schema":"https://wellknown.network/schemas/agent-record-v1.json","schemaVersion":"1","id":"ag_2w3u8phjyzj4","handle":"lightning-mcp","url":"https://wellknown.network/agents/lightning-mcp","links":{"self":"https://wellknown.network/agents/lightning-mcp/record.json","html":"https://wellknown.network/agents/lightning-mcp","markdown":"https://wellknown.network/agents/lightning-mcp/record.md","api":"https://wellknown.network/api/v1/agents/lightning-mcp","status":"https://wellknown.network/api/v1/agents/lightning-mcp/status","claim":"https://wellknown.network/agents/lightning-mcp/claim","claimApi":"https://wellknown.network/api/v1/claims","claimDescriptor":"https://wellknown.network/agents/lightning-mcp/claim.json","badge":"https://wellknown.network/agents/lightning-mcp/badge.svg","openapi":"https://wellknown.network/openapi.json","history":"https://wellknown.network/api/v1/agents/lightning-mcp/history","tools":"https://wellknown.network/api/v1/agents/lightning-mcp/tools"},"ard":{"identifier":"urn:air::server:lightning-mcp","type":"application/mcp-server-card+json"},"kind":"mcp_server","declared":{"name":"lightning-mcp","summary":"Model Context Protocol (MCP) server for PyTorch Lightning","description":"# PyTorch Lightning MCP (Model Context Protocol)\n\nAn integration layer that exposes **PyTorch Lightning** through a structured, machine-readable API.\n\nIntended for programmatic use by tools, agents, and orchestration systems.\n\n## Features\n\n* Structured training and inspection APIs\n* Real PyTorch Lightning execution\n* Explicit, config-driven behavior\n* Safe model instantiation\n* Stdio and HTTP servers\n* Fully tested core logic\n* Clean separation between protocol, capabilities, and transport\n\n## Project Structure\n\n```\nsrc/lightning_mcp/\n├── protocol.py          # Request / response schema\n├── handlers/\n│   ├── train.py         # Training capability\n│   └── inspect.py       # Inspection capability\n├── lightning/\n│   └── trainer.py       # Lightning integration boundary\n├── server.py            # Stdio server\n├── http_server.py       # HTTP server (FastAPI)\n├── models/\n│   └── simple.py        # Example LightningModule\n├── tools.py             # Expose tools\ntests/                   # Simple test suite\n```\n\n## Requirements\n\n* Python 3.10 – 3.12\n* PyTorch Lightning (compatible versions)\n* uv (recommended)\n\n## Installation (using uv)\n\n### 1. Install uv (if not already installed)\n\n```bash\ncurl -Ls https://astral.sh/uv/install.sh | sh\n```\n\nRestart your shell after installation.\n\nVerify:\n\n```bash\nuv --version\n```\n\n### 2. Clone the repository\n\n```bash\ngit clone https://github.com/<your-org>/lightning-mcp.git\ncd lightning-mcp\n```\n\n### 3. Install dependencies\n\nTo install all dependencies (including server extras):\n\n```bash\nuv sync --all-extras\n```\n\nThis will:\n\n* create a local virtual environment\n* install PyTorch Lightning and dependencies\n* install HTTP server dependencies (FastAPI, Uvicorn)\n\nNo manual venv management is required.\n\n## Usage\n\n### Training (in-process)\n\n```python\nfrom lightning_mcp.handlers.train import TrainHandler\nfrom lightning_mcp.protocol import MCPRequest\n\nhandler = TrainHandler()\n\nrequest = MCPRequest(\n    id=\"train-1\",\n    method=\"lightning.train\",\n   …","publisher":{"name":"Pratyush Mishra","url":null},"homepage":null,"repository":null,"version":"0.5.0","license":"Apache-2.0","protocols":["mcp"],"tags":["mcp"],"pricing":null,"endpoints":[{"url":"pypi:lightning-mcp","type":"package_pypi","auth":null,"probeable":false}],"skills":null,"tools":null,"extra":null,"attribution":{"kind":"pypi","name":"pypi","license":"pypi","summary":"pypi","version":"pypi","description":"pypi","publisherName":"pypi"}},"derived":{"capabilities":[{"slug":"dev.version-control","name":"Version Control","confidence":1,"provenance":"derived"},{"slug":"dev.terminal","name":"Terminal & Shell","confidence":0.745,"provenance":"derived"}],"categories":["dev"],"language":"en"},"observed":{"status":"unknown","statusReason":"Distributed as a package to run locally; no network endpoint to check.","lastOkAt":null,"lastProbedAt":null,"statusComputedAt":null,"reliability30d":null,"latestObservations":[],"tools":null,"package":{"name":"lightning-mcp","registry":"pypi","observedAt":"2026-10-28T08:28:19.550Z","publishedAt":"2026-01-04T14:27:18.799540Z","latestVersion":"0.5.0"},"toolSurface":null,"endpointFacts":[]},"verification":{"claimed":false,"claimedAt":null,"proofs":[]},"provenance":{"sources":[{"source":"pypi","key":"lightning-mcp","url":"https://pypi.org/project/lightning-mcp/","firstSeenAt":"2026-09-09T23:25:28.021Z","fetchedAt":"2026-10-28T08:27:51.923Z","normalizedAt":"2026-10-28T08:27:51.923Z"}]},"firstSeenAt":"2026-09-09T23:25:28.021Z","updatedAt":"2026-10-28T08:28:19.550Z"}