Skip to main content
DevelopersAPIs & Core SDKs

Deep Integration with OpenCSG Ecosystem

Integrate model weights, fine-tuning, inference, and agents via standard REST APIs, Python/Go SDKs, and Git.

Learning Roadmap & Milestones

Complete the following steps sequentially. Each milestone delivers concrete, verifiable outcomes.

  1. 1

    Setup Environment & API Credentials

    Generate Personal Access Tokens and configure auth headers for local development.

    Developer Quickstart
  2. 2

    Git & CLI Asset Operations

    Manage large model files and dataset checkpoints via Git LFS and CSGHub CLI.

    CLI Operations Guide
  3. 3

    Deploy & Call Model Inference API

    Launch OpenAI-compatible online inference endpoints with streaming response support.

    Inference API Reference
  4. 4

    Integrate Official Python / Go SDK

    Interact with repository assets and inference models programmatically using official SDKs.

    SDK Quickstart Guide
  5. 5

    Build Custom MCP Servers & Tools

    Expose internal databases and private APIs to AI agents following the MCP standard.

    Develop MCP Extensions
  6. 6

    Publish & Share Model Assets

    Package Model Cards, release tagged versions, and collaborate with team members.

    Publish Assets

Recommended Hands-on Tasks

Turn theoretical knowledge into verifiable hands-on practice.

Deploy Model Inference & Verify APIsLaunch an online LLM inference endpoint and verify OpenAI-compatible completions via cURL.
30–40 minsIntermediateAPI & SDK
Create Model & Dataset RepositoriesManage large model weights and datasets with Git LFS and native versioning on CSGHub.
15–20 minsBeginnerCSGHub
Integrate MCP Tools with AI AgentsExtend agent capabilities with external databases and tools using the open Model Context Protocol (MCP).
25–35 minsIntermediateAgenticHub