AI Link Lab

Skills

lark-cli: The Official Lark CLI That Puts 2500+ APIs in Your Terminal

lark-cli is the official Lark/Feishu CLI tool by the larksuite team — 200+ curated commands, 26 Agent Skills, three-layer command system. Here's what it does, why it matters, and how to get started.

3 hours ago

Matt Pocock's Skills: 122k Stars and a TDD Method That Actually Works

Matt Pocock's Skills project has 222k installs and 122k stars. The TDD skill uses vertical slices, integration tests, and red-green-refactor cycles to stop AI agents from writing buggy, mock-spaghetti code. A deep dive into how it works.

6 hours ago

Anthropic Frontend-Design Skill: Stop AI-Generated UI From Looking the Same

Anthropic's frontend-design skill makes AI coding agents generate bold, distinctive interfaces instead of generic purple-on-white UI. Typography, color, motion, layout — all intentionally designed, never default.

2 days ago

Grill-with-Docs: Make Your AI Agent Question Your Plan Before Writing Code

Grill-with-docs is Matt Pocock's AI Agent skill that stress-tests your plans with relentless questioning before a single line of code. Auto-updates CONTEXT.md and ADRs to build a shared language between you and your AI agent. Why the #1 skill from a 120k-star repo is 'ask first, code later.'

2 days ago

Obsidian Skills: teach agents to actually work with your knowledge base

A 35k-star open-source project that teaches AI agents to use Obsidian Markdown, Bases, JSON Canvas, and CLI. Standard Agent Skill format compatible with Claude Code, Codex, OpenCode.

2 days ago

Remotion Best Practices: 30 Rules That Teach AI Agents How to Build Videos in React

Remotion's official agent skill installs 30+ specialized rules for video development — covering animations, audio, captions, transitions, and FFmpeg. It turns any AI coding session into a Remotion project that actually renders correctly.

2 days ago

Agents

DeepSeek GUI: A Desktop Agent Workspace for DeepSeek Models

An Electron-based desktop workspace for DeepSeek models. Code mode, Write mode, IM integration, and a cache-first agent runtime called Kun that maximizes token ROI.

4 days ago

OpenChamber: When Your AI Coding Agent Needs More Than a Terminal

OpenChamber is an open-source desktop and web interface for OpenCode. It provides a branchable chat timeline, worktree-isolated multi-agent runs, Cloudflare tunnel remote access, inline diff review, and Git/GitHub workflows — all around your terminal AI agent.

2026-05-30

Multica: When Coding Agents Multiply, Terminals Stop Scaling

Multica is an open-source Human + Agent project management platform that puts Claude Code, Codex, Cursor Agent, and other coding agents into one board, task lifecycle, skill library, and runtime panel.

2026-05-20

Vibe-Trading: Turn Trading Ideas Into Backtestable Agent Research

Vibe-Trading is an open-source finance research agent workspace. It connects natural-language questions to market data, strategy generation, backtesting, reports, persistent memory, multi-agent teams, and MCP integrations. It is built for research, simulation, and backtesting, not live trade execution.

2026-05-18

Five Things Claude Code 102 Teaches Academic Researchers

Mushtaq Bilal's second Claude Code tutorial went viral with 440K views. Behind the numbers is a clear signal: AI agents for research are moving from experiment to standard practice.

2026-05-13

The Agent Harness: Why Your Model Isn't the Problem

LangChain jumped 25 spots on TerminalBench without touching the model. The harness — not the LLM — is what separates demos from production. Here are the 12 components that matter.

2026-05-11

Development

Codex++: Unlock Plugins, Delete Sessions, and Route API Traffic for Codex App

Codex App locks plugins when using API Key login, won't let you delete sessions, and can't use custom API proxies. Codex++ fixes these through CDP injection without touching the original install.

4 days ago

Spend One Hour Setting Up Claude Code Harness — See the Difference Immediately

Same model, same Claude Code — with or without a Harness setup, the output gap is staggering. Build a compounding agent configuration system in five steps: CLAUDE.md, Hooks, Skills, LSP, and Sub-agents.

2026-05-28

Stop Writing Markdown — Use HTML with Claude Code

Thariq, a Claude Code engineer at Anthropic, explains why he's replacing Markdown with HTML as the primary output format for AI agents — and how you can too.

2026-05-10

OpenSpec: agree with your AI before you write a single line

AI coding without specs means vague prompts and unpredictable results. OpenSpec brings predictability without the ceremony.

2026-05-07

Open Source

IKEA 3D Model Downloader: The Browser Extension IKEA Should Have Built

IKEA's website has beautiful 3D models for thousands of products but you cannot download them. This open-source Tampermonkey script changes that. A deep look at a browser tool that turns one specific friction point into a button press.

2 days ago

Cell Architecture Studio: Explore Biology in 3D, Open Source

An interactive 3D cell biology gallery built with React and Three.js. Seven cell types, GLB models, AI Tutor, comparison mode — 1.3k stars and open source.

4 days ago

LongCat-Video: Meituan's Open-Source 13.6B Video Generation Model — Long Video Is the Real Battleground

Most video generation models fall apart after 5 seconds — color drift, temporal discontinuity, quality degradation. LongCat-Video solves the core challenge of long video generation through native Video-Continuation pretraining, unifying T2V, I2V, and video continuation in a single 13.6B model at commercial-grade quality.

2026-05-31

Claw Code: Why AI Programming Needs Open-Source Agent Harnesses

Claw Code isn't just another open-source Claude Code clone. It represents a paradigm shift: the center of gravity in AI programming is moving from model capability to Agent Harness capability. Here's why that matters.

2026-05-29

skills.sh: Vercel Is Building the npm for Agent Skills

skills.sh is Vercel's open-source CLI and directory for the Agent Skills ecosystem. One command to install reusable skills across 50+ agents, with 410K+ total installs. A deep look at its architecture, cross-agent compatibility, and ecosystem momentum.

2026-05-26

AntV Infographic Turns AI Output Into Editable Visuals

AntV Infographic is an open-source declarative infographic engine from AntV. This review covers its AI-friendly syntax, streaming SVG rendering, templates, skills, and trade-offs.

2026-05-23

Cool Sites

mesh3d: A Better Place to Find 3D Website Inspiration

mesh3d is a curated gallery of Three.js, WebGL, and interactive 3D websites. Here is why designers, developers, and creative teams should bookmark it.

2026-05-23

Google Labs: Google's Secret Stash of Playable AI Experiments

I stumbled into labs.google and discovered a dozen experimental AI tools Google is quietly incubating — from the paper-to-podcast NotebookLM to the prompt-free Whisk. Each one is worth an afternoon.

2026-05-23

Tutorials

Obsidian Beginner Guide: Install It and Understand the Interface

A beginner-friendly Obsidian tutorial that covers installation, vaults, the file pane, editor, right sidebar, and basic settings so you can start writing on day one.

6 days ago

Obsidian Tutorial: What Is Obsidian and Why Is It Called a Second Brain Tool?

A beginner-friendly guide to Obsidian, vaults, note links, backlinks, graph view, plugins, and a simple first workflow for building a personal knowledge base.

7 days ago