A Claude Code plugin that provides expert guidance for the OpenGradient Python SDK - a decentralized AI inference platform with verified execution and on-chain settlement.
OpenGradient is a decentralized AI inference network that runs models inside Trusted Execution Environments (TEEs). Every inference call produces a cryptographic proof of correct execution, and payment is settled on-chain via the x402 protocol on Base Sepolia. This means you can call models from OpenAI, Anthropic, Google, and xAI through a single unified API — with verifiable, tamper-proof results.
When installed, this plugin adds the /opengradient-sdk skill to Claude Code. It gives Claude deep knowledge of the OpenGradient Python SDK so it can help you write correct, idiomatic code for:
- Verified LLM inference — Chat completions via TEE with cryptographic attestation
- x402 payment settlement — On-chain receipts on Base Sepolia (
SETTLE,SETTLE_METADATA,SETTLE_BATCH) - Multi-provider models — Access OpenAI, Anthropic, Google, and xAI models through one API
- Streaming & tool calling — Full multi-turn agent loops with function definitions
- On-chain ONNX model inference (alpha) — Run ONNX models directly on-chain
- LangChain integration — Build ReAct agents using the OpenGradient adapter
- Digital twins chat — Specialized chat API for digital twin interactions
- Model Hub — Upload, manage, and query models
- Claude Code installed and running
- A Base Sepolia wallet funded with OPG tokens (required for LLM inference)
claude plugin marketplace add https://github.com/OpenGradient/claude-pluginsThis registers the OpenGradient marketplace and installs the plugin, making the /opengradient-sdk skill available in your Claude Code sessions.
Invoke the skill directly:
/opengradient-sdk write a streaming chat completion with tool calling
Or just describe what you want to build with OpenGradient - Claude will automatically use the skill when relevant.
/opengradient-sdk create a multi-turn agent that uses tool calling to fetch weather data
/opengradient-sdk build a LangChain ReAct agent with OpenGradient
/opengradient-sdk upload an ONNX model to the model hub
/opengradient-sdk show me how to use x402 settlement with metadata
