Art & Design

Explore curated resources for harness engineering to build reliable AI agent systems with context, evals, observability, and runtime controls

Recover Claude Code and reset Python state for a clean CLI port with preserved module logic and no UI layer

Recover Claude Code source for TypeScript and React UI analysis and reuse

Build, modify, and run a research fork of Claude Code from the leaked TypeScript source snapshot

Restore, build, and debug the Claude Code 2.1.88 source tree for continued development and patching

Explore a Claude Code 2.1.88 source snapshot with CLI analysis, package files, and buddy TUI and web demos for code study.

Recover and reset Claude Code in Python with a faithful CLI port that preserves core logic and removes the UI layer

Organize Claude Code docs in Korean and English for fast access to guides, analysis, and best practices

Build a TypeScript CLI POC for AI-assisted coding with Claude API, tool calling, memory, and task workflow support

Build a token-saving AI coding assistant for OpenAI and Chinese LLMs with AST-based context, self-healing, and intent checks

Rebuilt Claude Code CLI in TypeScript for full local use, with React + Ink UI and unlocked slash commands

Proxy Claude Code-style OAuth and Messages API with token management and streaming support built on Bun and Hono

Build an unofficial Claude Code CLI from source for research, Bun builds, and Node.js use with custom API endpoints

Review diffs, files, and docs in a TUI with inline notes, then export structured annotations for scripts and AI tools

Explore 10 Claude Code prompt engineering patterns with Python examples and practical guidance for building better AI workflows

Build coding agents with a clean-room harness for streaming loops, tool control, session save, secure creds, and layered config

A Claude skill that orchestrates a 5-round, 7-persona product team review (eng / UX / UI / creative / growth / data / CoS). Personas cross-pollinate, harden each round, escalate to the user only on deadlock.

Build production-grade Claude prompts with rules from Claude Code source for structured, high-quality LLM prompt generation

Turn academic notes, meetings, chats, and emails into a reusable AI supervisor skill for clear, actionable guidance and stable research workflows

Convert raw documents and notes into an interlinked markdown wiki to build a persistent, browsable knowledge base.

Build functional AI agents with tool-use capabilities using this simple, step-by-step framework for developers.

Explore LLM prompts by building parallel implementation paths in git worktrees to resolve decision points through user input.

Integrate Claude with Obsidian to build a persistent, searchable knowledge vault that automatically connects your notes and sources.

Learn to build and control AI agents with this practical guide to the Hermes Agent framework.