subagent-driven-development

An agentic skills framework & software development methodology that works.

⚡ Modular prompt-engineering framework for AI Agents (Cursor, Claude Code, Copilot). Forces structured SDLC, linear execution, context isolation, and systematic debugging via a standalone Windows launcher. Stop vibe coding, start agentic discipline! 🚀

A warn-only runtime layer for Claude Code agents. Keeps feature state, scope and handoffs on disk so they survive compaction, and requires an evidence path before a done claim. Never blocks a tool call. Companion to superpowers; bash + git, no MCP.

A lightweight agents and skills set to handle any kind of task! Designed to create a professional workflow with dynamic team assembly of specialized agents tailored to each task, parallel execution, optimized context windows, and a self-learning cycle.

Self-contained agentic development harness for Claude Code — one command runs the full feature pipeline (plan → execute → double review) with security gates baked in.

Beyond vibes, into louke (craft)

Turn one feature request into a coordinated Codex delivery workflow: requirements, implementation, independent QC, and human acceptance.

Open‑source AI agent skills for Claude Code and Codex, covering video creation, short‑form production, downloading, Xiaohongshu content, and AI content workflows.

Compile natural language specs into local neural functions for text classification, extraction, and fuzzy matching tasks in your AI coding agent.

Safety-hardened fork of obra/superpowers. Same skills library, plus a mandatory safety-check preflight (5 hard gates) before any skill runs. Defends computer, hardware, budget, and data.

Architecture-aware autonomous software engineering with automatic local experience learning

Manage reversible themes for Chromium and Electron AI desktop apps using installable agent skills.

Claude Code skill: delegate implementation to a secondary code-agent CLI, monitor for stalls, and independently review every diff — with the field lessons that earned each rule

First-person agent skills: every claim carries a verdict backed by evidence produced this turn.

一套轻量化，可复用的 Codex 子智能体配置与编排约定，通过读写分离、 Sandbox 隔离、任务契约、有限重试和结构化交付，使多智能体协作 更加安全、稳定并且可审查。

Project-driven agentic engineering: an MCP tool surface and a skill that let an agent pick the next item across many repositories, work it, and close it with a pull request.

Guide me, agent

An agentic skills framework & software development methodology that works.

Install production-ready Claude Code skills for better prompts, React types, and automated agentic workflows.