Coding & Development

Explore a LangChain document pipeline for RAG, from ingestion and embeddings to retrieval and LLM response generation

Display Claude Code usage on a BLE tri-color e-ink screen for always-visible limits and session stats

Catalog real-world AI agent incidents, attack techniques, CVEs, and defensive tools for secure autonomous agent defense

ACP is a reusable democratic coordination primitive for discussion under load. Relay is the reference implementation used to validate that primitive in practice.

Levers for AI coding agents — multiply the agent's force, divide the human's effort.

Genetic exploration of solution spaces for AI coding agents — gather a verifiable objective, evolve variants, apply the winner.

Command-line interface for Twist

一个零配置的 Prompt 优化 Skill：通过追问澄清与结构化重写，让模糊的 AI Agent 指令变得清晰、可控、可执行。

This is my personal configuration of skills as a Ruby on Rails Dev

iEvo — self-evolving plugin for Claude Code. Capture lessons, patch local agents and skills, replay logs on upstream updates.

Governed operator skill for structured, scoped, validated and traceable agent work with contracts, event logging and controlled mini-prompt chains.

Install KDNA skills for any AI agent: Claude Code, Codex, OpenCode, Cursor, and GitHub Copilot. One skill, one format, multiple agents.

Agent skills for implementing Malaysia payment gateway integrations.

CLI tool for managing software specifications as a typed graph

A collection of skills I use regularly

⛏️ Document the content AI Coding leaves behind. — 记录AI协作中的意图矿脉

51 opinionated skills for AI coding agents. 27 ship with AST/shell verifiers. -72% verifier-visible slop on Sonnet 4.6, -58% on Haiku 4.5 (20 adversarial prompts, 3 runs/arm).

Manager for AI skills, subagents, and hooks - install and sync across AI clients & IDEs

VMware vCenter/ESXi AI-powered monitoring and operations. Two skills: vmware-monitor (read-only, safe) and vmware-aiops (full operations) | Claude Code Skill

AVI (NSX Advanced Load Balancer) management and AKO Kubernetes operations tool

Lionclaw is a local control plane for AI coding agents. It runs AI agents as durable, auditable workers with explicit state, skills, channels, schedules, and bounded permissions.

Laravel conventions, as agent skills. Your team's rulebook — portable across Claude Code, Codex, Cursor, and anything else SKILL.md.

Light-weight AI agent evaluation and optimization framework

Terminal-native File Agent Platform