Coding & Development

Define an AI agent once, generate the config every tool expects (CLAUDE.md, AGENTS.md, GEMINI.md, .cursor/rules). Clean-room, built on open-gitagent (MIT).

AutoML for Claude Code — multi-agent skill that designs, runs, and reports ML experiments from a text prompt

Skill that designs paste-ready /goal prompts for Claude Code by enforcing the Three Pillars and the 4-Element Structure. Works in Cursor, Claude Code, and Codex.

Build agentic AI workflows with curated tools, MCP server reviews, and technical recipes for software developers.

R-Pay: a fictional UPI-style payment sandbox for India, demonstrating Claude Code agent workflows — state machines, idempotency, incident response, subagents, and RCA — through a simulated midnight payment outage.

Audit AI coding agents, separate DNA from tech stack skills, and embed reusable methodology into agent definitions

Build production-ready AI agent configs for Claude Code with structured workflows, special agents, and clear rules

Orchestrate AI agents with schedules, approval gates, budget limits, and live monitoring for Claude and GPT teams

Compare code reviews across models in Claude Code with ranked todos and subagent dispatch via MCP.

Repo containing backup of all my personal Agentic Coding Helpers - Skills, Custom Agents, Custom Instructions and everything else

Connect AI agents to your real Chrome browser with shared sessions, cookies, and logins via a Chrome extension and MCP server

Autonomous AI farming agent — Gemini + Arize MCP. Hackathon: Building Agents for Real-World Challenges.

Self-hosted Docker stack for cross-platform AI agent rooms (Claude Code + ChatGPT + any LLM). MIT.

MCP ToDoList for plan and loop for LLM-powered agents. Built with Python and FastMCP. Provides tools for sequential task management with built-in failure detection and recovery

MCP server for INDmoney integrations — enabling AI agents to access Indian stocks, portfolios, mutual funds, and financial workflows through Python and the Model Context Protocol.

SSH-friendly headless browser MCP server for AI agents.

Official Prysmid MCP server — manage workspaces, apps OIDC, login policy, IdPs, branding, users from Claude Code/Cursor/any MCP-compatible agent

Pilot Protocol skill injector — reconciler that installs SKILL.md into agent-tool dirs (Claude Code, OpenClaw, PicoClaw, OpenHands, Hermes)

Part 2 of the AI and agent skills collection, with 650+ skill folders focused on reusable workflows, security playbooks, cloud implementation guides, scripts, references, and assets for builders and operators.

Agent Vibe Coding skills for AI coding agents: Pi harness patterns and Craft Agents app architecture.

pi-sdk-agent qclaw is vibe code agent, use pi sdk , do like coding agent, openclaw, codex, claude code, and more...

Grok Build resources, reusable .grok/skills, AGENTS.md templates, hooks, prompts, and starter workflows.

CI AutoHealing monitors [Konflux](https://konflux-ci.dev/) CI/CD pipelines, detects build failures and policy violations, uses LLMs to diagnose root causes, and autonomously generates fix PRs — all orchestrated through a kubectl-style CLI, an MCP server for AI agents, and a REST API.

AGENTS.md (身分定義) , Rules (規範約束) , Workflows (工作流程) , SKILL.md (技能工具描述)