A collection of reusable skills for Claude Code and compatible AI agents. Each skill is a folder with a SKILL.md file that tells the agent how to handle a specific task.
🗂️ Browse skills
| Folder | What's inside |
|---|
| 🧩 Core | Reusable skills that work across repositories and install into your agent's user-level skills directory. |
| 🧭 Domain | Field-specific skills (engineering, security, business, media, API, science) mirrored from the org's agent workspaces. Portable, installed on request. |
| 🗂️ Projects | Additional skills for specific repositories, installed into the project they belong to. |
| 🖥️ Showcase site | A static page that presents every Core skill as a card, with screenshots and install commands. Published at agent-skills.ai-automation-tools.dev. |
📦 Core skills
| Skill | Category | What it does |
|---|
| ✉️ html-email-templates | Automation | Creates HTML emails for notifications, reports, and digests. |
| 🏭 automated-dev-pipeline | Automation | Sets up scheduled AI agents that build projects through PRs, publish them to a GitHub account or organization behind scripted gates and a human approval, and keep every repo current. Also covers pipelines whose stages are folders in one repo that deploys a site. |
| 🧠 agent-workspace-builder | Automation | Creates or standardizes a specialized agent workspace repo (AGENTS.md doctrine, docs tree, capability reference, curated resources, portable skills) and sets up its weekly roadmap and upkeep automations. |
| 🚦 task-router | Automation | Chooses a workflow and model tier based on the task. |
| 📈 business-plan-builder | Business | Evaluates business ideas and writes a decision memo and build plan. |
|
🧭 Domain skills
Field-specific skills from the org's five agent workspaces, copied here unchanged by a weekly
harvest. They're kept apart from Core so Core stays a short list, and they install only
when you ask: install-skills.ps1 -Domain. Edit them in their workspace repo; the
harvest ledger says which one.
| Domain folder | What's inside |
|---|
| 🛠️ Engineering | How to build software well: debugging, performance work, and test strategy. From fullstack-agent. |
| 🌐 API | Working with third-party APIs: judging one before you integrate, writing clients that survive failures, staying inside rate limits and quotas, normalizing responses into one traceable shape, and testing integrations without hitting live services. From api-agent. |
| 📈 Business | Field tactics for business decisions: running customer discovery interviews, planning go-to-market, and setting prices. From business-agent. |
| 🔐 Security | Communicating and steering security work: turning verified findings into a report a client acts on, disclosing a vulnerability responsibly, and measuring a security program with metrics, SLAs and a risk register. From security-agent. |
| 🎬 Media | Getting a media asset ready to publish and learning from it afterward: captions, alt text, accessibility checks, rights clearance, and reading post-publish analytics. From media-agent. |
| 🔬 Science | Doing and judging research: appraising a paper's design, bias and statistics, managing research data so it stays findable and reusable, and making figures that show the data honestly. From . |
🗂️ Project skills
Each folder lists the project's existing skills and any additional skills provided here.
| Project folder | Additional skills in this repo |
|---|
| 💬 agent-chat | agent-chat-conv-types, agent-chat-delivery and agent-chat-run-triage. |
| ⏱️ cronsole | cronsole-windows-jobs and cronsole-claude-routines. |
| 📡 edge-radar | edge-radar-strategy-profiles and edge-radar-strategy-evidence. |
| 📉 edge-spectrum | edge-spectrum-hub-tool, edge-spectrum-dataset and edge-spectrum-endpoint. |
| 🌐 project-hub | project-hub-scan, project-hub-client and project-hub-demo. (project-hub-scaffold is listed under Core.) |
See the Using Skills guide for installation and authoring instructions, or browse Resources for examples and supporting material. What's planned next is in the roadmap, and what has already changed is in the changelog.
