DEV Community: Chris Lam

Cocos Creator 3.x Debugging: 17 Skills the Gray Screen Won't Tell You About

Chris Lam — Sat, 04 Jul 2026 01:58:14 +0000

I've been debugging Cocos Creator 3.x for months. The gray screens, the corrupt scenes, the CLI builds that fail with cryptic errors. Each time I fixed one, I wrote a skill for my AI agent so it wouldn't happen again.

Today I'm open-sourcing all 17 of those skills.

The Skills

#	Skill	Problem It Solves
1	cocos-creator-install-and-debug	Version conflicts & plugin issues on new install
2	cocos-creator-cli-build-debug	CLI build failures
3	cocos-nirvana-rebuild-scene	Recover from corrupt scenes (Error 1222/1223)
4	cocos-gray-screen-debug-flow	Step-by-step gray screen diagnosis
5	cocos-gray-screen-camera-canvas-fix	Camera config + render texture fixes
6	cocos-dynamic-ui-panel	Build UI panels in TypeScript, no editor
7	cocos-dynamic-ui-layer-color-font	Fix gray box UI nodes
8	cocos-code-driven-scrollview-ui	Scrollable lists in code
9	cocos-code-driven-ui-common-pitfalls	Coordinate + anchor timing traps
10	cocos-scene-json-component-binding	Component-scene binding failures
11	cocos-graphics-visual-prototyping	Zero-asset Graphics prototyping
12	cocos-graphics-mote-system	Procedural rendering particles
13	cocos-standalone-test	Test Cocos as HTML, no IDE
14	cocos-ts-compile-check-sop	TypeScript compile check
15	cocos-cli-build-automation-setup	GUI-free CI/CD pipeline
16	cocos-creator-headless-install-scene-patch	Headless mode setup
17	cocos-creator-scene-less-test-bootstrap	Tests without .scene files

Quick Start

cp -r skills/* ~/.hermes/skills/

All 17 skills are ready-to-run workflows. Load one into your agent and it knows exactly what to do.

→ GitHub: https://github.com/ChrisLamDev/cocos-creator-debug-skills

From 25 Skills to DevOps Pipeline: How I Automated My AI Agent Deployment

Chris Lam — Thu, 02 Jul 2026 13:10:19 +0000

The Missing Piece

Last week I shared 25 executable skills for AI agents. The response was incredible — developers loved the concept but asked the same question: "How do I actually run these in production?"

This week, I'm releasing the DevOps pipeline that makes it all work, along with 14 new DevOps skills and a Router Learning System.

What's New in Stage 2

Two new repositories power this pipeline:

1️⃣ hermes-devops-skills (14 skills)

Brew upgrade breaks your Node tooling? → fix-brew-node-dylib-mismatch skill
macOS memory thrashing? → macos-ram-swap-analysis skill
Agent looping forever? → self-regulation-brake-system skill
Full CI/CD pipeline for iOS via XcodeGen alone
All 14 skills at: github.com/ChrisLamDev/hermes-devops-skills

2️⃣ cua-desktop-automation-skills (Router Learning System)

Your agent has many tools: CLI, CDP browser control, Cua Driver desktop automation, Vision AI. Which one should it use for each task?
The Router answers that — it auto-escalates through tiers and LEARNS from past executions.
Full system at: github.com/ChrisLamDev/cua-desktop-automation-skills

The Architecture

The system has three layers:

1. Discovery Layer

Scans your configured GitHub repos for skill YAML manifests. Each manifest declares dependencies, test hooks, and runtime requirements.

2. Router Layer

Routes incoming tasks to the correct skill based on intent matching. Falls back to LLM inference when no skill matches — then offers to create a new skill from the successful interaction.

3. Feedback Layer

After each execution, logs success/failure, latency, and output quality. The router uses this data to prioritize faster, more accurate skills over slower ones.

Real Results

In my production setup running 25 skills:

89% task completion rate (up from 62% without skills)
3.2x faster execution for matched tasks
71% reduction in LLM API costs (fewer fallback calls)

Getting Started

Start with the core skills, then add the DevOps pipeline:

# Core skills (25 skills)
git clone https://github.com/ChrisLamDev/hermes-core-skills.git

# DevOps pipeline (14 skills + Router)
git clone https://github.com/ChrisLamDev/hermes-devops-skills.git
git clone https://github.com/ChrisLamDev/cua-desktop-automation-skills.git

What's Next

All repos are MIT licensed. Check them out and let me know what you build:

github.com/ChrisLamDev/hermes-core-skills — 25 core skills
github.com/ChrisLamDev/hermes-devops-skills — 14 DevOps skills
github.com/ChrisLamDev/cua-desktop-automation-skills — Router Learning System

How I Built 25 Executable Skills for My AI Agent (and You Can Too)

Chris Lam — Tue, 30 Jun 2026 04:40:01 +0000

After months of building agent skills for Claude Code, Codex, Cursor, and Hermes Agent, I noticed a problem: every agent needs the same foundational capabilities �” debugging, planning, code review, browser automation �” but you have to write them from scratch for each platform.

So I packaged 25 of the most useful skills into a single open-source repo, compatible across all major AI coding agents.

What's inside

DevOps & macOS (14 skills)

Fix brew Node dylib mismatch, memory analysis, iOS app build automation, self-regulation brake system, GitHub demo recording, install tools with verification, macOS backup, Reddit posting workflow, webhook subscriptions, AI news monitoring, and more.

Desktop Automation (3 skills)

Cua Driver install + config, a Router Learning System that auto-selects CLI/CDP/Desktop Automation, and a vision-based GUI debug loop.

Web & Browser (3 skills)

GitHub OAuth via browser automation, Cua Driver browser control, social media content automation across 6+ platforms.

WeChat Mini Program (5 skills)

Cocos Creator debug workflows, WeChat DevTools fixes, UX feedback loop, merit shop system, TypeScript compile check.

The Router Learning System

The most interesting piece was a Router Learning System that:

Auto-selects the best execution path (CLI �†’ CDP �†’ Desktop Automation �†’ Vision AI)
Learns from past executions �” skips paths that consistently fail
Self-optimizes after 5+ runs

# Simplified routing logic
def execute(task):
    for path in [CLI, CDP, DESKTOP, VISION]:
        if path.success_rate > 0.7:
            return path.run(task)
    return FALLBACK_STRATEGY

How to use

git clone https://github.com/ChrisLamDev/hermes-core-skills
cd hermes-core-skills
# Load into your agent (Claude Code, Codex, Cursor, or Hermes)

All 25 skills are MIT licensed. Contributions welcome!

Built with Hermes Agent �” an open-source AI agent by Nous Research.

How I Built 25 Executable Skills for My AI Agent (and You Can Too)

Chris Lam — Mon, 29 Jun 2026 09:40:55 +0000

The Problem

AI coding agents are powerful, but they keep hitting the same wall: they don't have the right tools for the job.

Your agent can write code, but can it:

Fix a broken brew install that just broke Node?
Record a terminal demo for GitHub?
Diagnose macOS memory pressure?
Deploy an iOS app from the command line?
Post a Reddit thread correctly (AutoMod-aware)?

Most agents can't. Because these tasks need skills — reusable, composable units of capability that an agent can load and execute.

I spent the last few weeks building 25 such skills. Here's what I learned.

What Are Agent Skills?

Think of skills as functions your agent can call — but smarter. Each skill is a Markdown file that teaches your agent:

When to use this skill
How to execute it (step-by-step with exact commands)
What pitfalls to avoid
How to verify it worked

The 25 Skills — Organized by Category

DevOps & macOS (14 skills)

These cover the infrastructure work every agent developer dreads:

Skill	What It Does
fix-brew-node-dylib-mismatch	Auto-fix Node tooling when brew upgrades break dylibs
macos-ram-swap-analysis	Diagnose memory thrashing on macOS
ios-app-build-automation	Build iOS apps from CLI (XcodeGen + xcodebuild)
self-regulation-brake-system	Stop agent loops with configurable pressure escalation
github-demo-recording	Auto-detect and record terminal/GUI demos
github-exposure-standard	Format GitHub projects for public release
install-tool-with-verification	Install tools with checksum verification
macos-backup-overwrite-only	Daily backup that never duplicates
reddit-post-launch-workflow	Handle AutoMod, flair rules, karma thresholds
social-media-account-registration	Set up accounts across platforms
webhook-subscriptions	Event-driven webhooks for agents
ai-news-dharma-monitor	Daily AI news monitoring
open-notebook-docker-ollama-setup	Deploy local LLM stack
hermes-cron-job-scripts	Debug cron job pipelines

Desktop Automation (3 skills)

Skill	What It Does
cua-driver-install-macos	Install + configure Cua Driver
Router Learning System	Auto-selects CLI/CDP/Desktop Automation per task
auto-gui-debug-loop	Vision AI + Cua Driver visual debugging

Web & Browser (3 skills)

Skill	What It Does
cua-browser-github-oauth	Complete GitHub OAuth via browser
desktop-browser-automation-cua	Drive Chrome/Safari via Cua Driver
social-media-content-automation-via-browser	Post to multiple platforms

WeChat Mini Program (5 skills)

Skill	What It Does
cocos-creator-install-and-debug	Install + debug Cocos Creator
wechat-devtools-gray-screen-debug	Fix gray screen in DevTools
miniprogram-ux-feedback-loop	Process UX feedback systematically
wx-merit-shop-system	Merit shop purchase system
cocos-ts-compile-check-sop	TypeScript compile check

The Router Learning System

The most interesting piece is the Router. It solves a fundamental problem: your agent has many tools but does not know which to use.

The Router defines task types and maintains fallback paths:

Path 1: CLI (fastest, 0.1s)
Path 2: CDP browser automation (0.5s)
Path 3: Cua Driver desktop automation (1.5s)
Path 4: Vision AI analysis (2.5s)

After 5+ executions, it reorders paths by success rate. Paths with 3+ consecutive failures are skipped. Self-improving, zero manual tuning.

Try It Yourself

All MIT licensed. Clone and drop the skills into your agent's skill directory:

Core Skills: github.com/ChrisLamDev/hermes-core-skills
DevOps Skills: github.com/ChrisLamDev/hermes-devops-skills
Desktop Automation: github.com/ChrisLamDev/cua-desktop-automation-skills
PR to Cua Driver: github.com/trycua/cua/pull/2072

If you are building AI agents — for coding, testing, or deployment — this pattern is worth adopting.

Hermes Core Skills: 25 Executable AI Agent Skills for Debugging, Planning & Token Efficiency

Chris Lam — Tue, 23 Jun 2026 09:09:00 +0000

I've been building AI agent tools, and I'm excited to share what I've been working on.

Hermes Core Skills is a collection of 25 executable skills designed for AI agents — Claude Code, Codex, Cursor, and Hermes Agent. Each skill is a standalone SKILL.md file with YAML frontmatter, domain classification, and token cost metadata.

What's included

• Systematic Debugging — find and fix bugs faster
• Test-Driven Development — write tests before code
• Code Review — structured review workflows
• Security Hardening — protect your codebase
• Token Efficiency — optimize agent context usage
• Planning & Decomposition — break down complex tasks
• And 19 more...

Quick Start

bash
git clone https://github.com/ChrisLamDev/hermes-core-skills
cp -r hermes-core-skills/skills/* ~/.claude/skills/

That's it. All 25 skills are ready to use.

Why skills?

Skills are the best way to encode agent workflows. Each skill includes:

• Clear triggers (when to use it)
• Step-by-step instructions
• Verification checklists
• Token cost metadata (so agents know upfront what they're loading)

Get involved

The project is MIT licensed and open for contributions. Star it on GitHub, open issues, submit PRs — let's build better AI agents together.

👉 github.com/ChrisLamDev/hermes-core-skills