Public AI agent skills by DannyMac180.
I write Attention Heads — deep, evidence-backed writing on AI, cognition, and agentic engineering. The Agentic Engineering Field Notes series is where I publish practical advice on the craft of using AI. Subscribe to get new posts to your inbox.
claude-mod-builder- Build, test, and ship a Claude Mod: a Claude Code plugin whose behaviour lives in a function-hooks module, hooking the engine's events as TypeScript functions($, e, next). Covers event and tier choice, drawing above the prompt, the plugin test kit, and the early-access gotchas.codex-dynamic-workflows- Plan and run supervised AI-agent dynamic workflows with goal mode, subagents or simulated work packets, approval gates, integration, verification, and reusable workflow artifacts.codex-relay- Relay: open one new native Codex task from your current task with a compact handoff, end your turn, and receive a single completion or blocker message back. Honors your chosen model and effort, or your default.explain-this- Explain any digital artifact (papers, articles, code) shaped by a persistent learner profile, with comprehension quizzes and spaced-repetition review. On first use it interviews you (~10 min) and creates~/.explain-this/.
If your AI agent supports skills, you can point it at the GitHub URL for a skill and ask it to install that skill:
Install the AI agent skill at https://github.com/DannyMac180/skills/tree/main/codex-dynamic-workflows
You can also clone this repo and copy a skill folder into your agent's skills directory. Adjust the destination path for your agent; this example uses Codex's default skills folder:
git clone https://github.com/DannyMac180/skills.git
cd skills
mkdir -p "${CODEX_HOME:-$HOME/.codex}/skills"
cp -R codex-dynamic-workflows "${CODEX_HOME:-$HOME/.codex}/skills/"Then start a new agent session and invoke it with:
Use $codex-dynamic-workflows to plan and run a supervised multi-agent workflow for this task: ...
MIT