shanraisshan/claude-code-best-practice

This page introduces an external open-source repository. It is not an HDATF product.

What it does

A collection of best practices and example implementations for Claude Code, covering subagents, commands, skills, hooks, MCP servers and orchestration workflows. Several concepts link to both a best-practice guide and an implementation file in the repository.

How it helps ATF

A reference for setting up Claude Code as an executor under Harness, including skills, hooks and subagents. Its orchestration workflow example is worth comparing with how Harness hands off a single task.

License

MIT Permissive. Commercial use and changes are allowed if the copyright notice is kept.

More in this category

• entireio/cli
  A CLI that hooks into a Git workflow to capture AI agent sessions, including prompts, transcripts, files touched and tool calls, and indexes them next to commits. Users can search that record and resume from earlier checkpoints.
• j178/prek
  A Git hook manager written in Rust that runs hooks before commits, on demand or in CI to format files, catch lint errors or detect secrets. It is a single binary compatible with existing pre-commit configurations and hooks.
• blackboardsh/electrobun
  A framework for building, updating and shipping compact cross-platform desktop apps in TypeScript. It includes Hutch, a native build and workspace CLI, and Cottontail, a JSC-based JavaScript runtime, with a platform layer in Zig, Objective-C and C++.
• MoonshotAI/kimi-cli
  Kimi CLI is a terminal AI agent that reads and edits code, runs shell commands, searches and fetches web pages, and plans and adjusts its actions. It is being replaced by Kimi Code CLI.
• millionco/react-doctor
  React Doctor deterministically scans React code for issues in state and effects, performance, architecture, security, accessibility and maintainability. It can run in CI, reporting only issues a pull request introduced, and can be installed as a skill for coding agents.

Only repositories in the ranked Trendshift lists are included, and the lists are used only to find candidates. We do not copy their ranks. Descriptions, licenses and star counts come from each GitHub repository. The notes are our own reading. We have not tested these projects, and a place on a trending list does not prove quality.