# hermes-devops-skills
<div align="center">
[](https://github.com/ChrisLamDev/hermes-devops-skills/stargazers)
[](https://github.com/ChrisLamDev/hermes-devops-skills/forks)
[](https://opensource.org/licenses/MIT)
[](skills/)
[](https://claude.ai)
[](https://openai.com)
[](https://cursor.sh)
[](https://github.com/NousResearch/hermes-agent)
[](https://star-history.com/#ChrisLamDev/hermes-devops-skills&Date)
</div>
**14 battle-tested macOS & GitHub DevOps skills for AI coding agents.**
## 📸 Demo
<div align="center">
<img src="screenshots/demo.svg" alt="hermes-devops-skills overview" width="100%">
<br><br>
<img src="screenshots/demo-real.png" alt="hermes-devops-skills live demo" width="90%" style="border-radius: 8px; border: 1px solid #30363d;">
<br>
<em>Live system diagnostics — macos-ram-swap-analysis probing total RAM, memory pages, and swap usage on macOS 26.5.1</em>
</div>
<br>
## 💡 Why These Skills?
AI agents waste **30% of their time reinventing infrastructure**. A `brew upgrade` silently breaks Node tooling. Diagnosing RAM pressure requires six terminals and a prayer. That first Reddit post? Shadowbanned before anyone sees it. These aren't one-off annoyances — they're recurring tax on every session, draining tokens and patience alike.
These **14 skills** encode hard-won DevOps patterns so your agent ships faster, debugs smarter, and stops making the same mistakes. Each skill is a self-contained playbook: load it once, apply it forever. No more wondering why the same `brew` fix worked last month but broke today — the pattern is captured, tested, and reproducible.
## You've been here before:
- 🔄 Brew upgrade broke your Node tooling again? `fix-brew-node-dylib-mismatch`
- 🎥 Need a demo GIF but hate manual recording? `github-demo-recording`
- 🚀 Releasing open-source without a playbook? `github-exposure-standard`
- 🔍 Finding the right GitHub project among 10,000 clones? `github-project-discovery-hybrid`
- 📦 Installing a new tool and forgetting to verify it works? `install-tool-with-verification`
- 📱 Building iOS apps from CLI without opening Xcode? `ios-app-build-automation`
- 💾 Backup overwriting instead of accumulating cruft? `macos-backup-overwrite-only`
- 🐍 Deploying a Python tool from GitHub on Apple Silicon? `macos-deploy-python-tool-from-github`
- 🧠 macOS RAM full — swap thrashing, what's eating your memory? `macos-ram-swap-analysis`
- 🐳 Docker + local LLM setup without 20 tabs open? `open-notebook-docker-ollama-setup`
- 📣 First Reddit post gets AutoMod-blocked? `reddit-post-launch-workflow`
- 🔁 Agent stuck in a loop burning tokens? `self-regulation-brake-system`
- 📝 Setting up social accounts across 5 platforms? `social-media-account-registration`
- 🔔 Need event-driven agent activation? `webhook-subscriptions`
## Quick Start
```bash
# Clone the repo
git clone https://github.com/ChrisLamDev/hermes-devops-skills.git
# Copy skills to your agent's skill directory
# For Hermes Agent:
cp -r skills/* ~/.hermes/skills/
# For Claude Code:
cp -r skills/* ~/.claude/skills/
# For Cursor:
# Add the skills/ directory in Cursor settings → Custom AI Agent → Skills Directory
```
## Highlight Skills
| Skill | Pain Point | One-liner |
|-------|-----------|-----------|
| `self-regulation-brake-system` | Agent loops endlessly, burning tokens | Auto-stops after 3 failures, prevents infinite retry |
| `github-exposure-standard` | 0-star repos nobody discovers | Release playbook with proven social media + README templates |
| `ios-app-build-automation` | Xcode GUI slow for CI | Full XcodeGen + xcodebuild pipeline, zero GUI |
| `macos-ram-swap-analysis` | "System application memory insufficient" | Identifies RAM shortage vs swap congestion in one command |
| `reddit-post-launch-workflow` | New accounts get shadowbanned | AutoMod-aware posting with karma-farming strategy |
| `github-demo-recording` | Manually recording demos is tedious | Auto-detect + record terminal/GUI demos for any repo |
## What's Inside
### 🐙 GitHub Automation (5)
- `github-demo-recording` — Auto-detect environment, record terminal/GUI demos
- `github-exposure-standard` — Release playbook for discoverable, trustworthy repos
- `github-project-discovery-hybrid` — Multi-angle GitHub project search & evaluation
- `reddit-post-launch-workflow` — Full Reddit launch with flair, karma, and AutoMod handling
- `webhook-subscriptions` — Event-driven agent activation via webhooks
### ⚙️ macOS Operations (5)
- `macos-backup-overwrite-only` — Daily backup that overwrites instead of accumulating
- `macos-deploy-python-tool-from-github` — Deploy Python tools on Apple Silicon
- `macos-ram-swap-analysis` — Diagnose memory pressure and swap congestion
- `install-tool-with-verification` — Install + verify + health-check in one command
- `fix-brew-node-dylib-mismatch` — Fix brew-upgrade broken Node browser tooling
### 🛠️ Developer Tooling (4)
- `ios-app-build-automation` — XcodeGen + xcodebuild CI pipeline (zero GUI)
- `open-notebook-docker-ollama-setup` — Open Notebook + local Ollama via Docker
- `self-regulation-brake-system` — Agent self-regulation with pressure escalation
- `social-media-account-registration` — Social account setup with content pre-generation
## Platform Compatibility
These skills work with any AI agent that supports skill/memory injection:
- **Hermes Agent** (native) — full support
- **Claude Code** — drop `skills/*` into `~/.claude/skills/`
- **Cursor** — add via Custom AI Agent settings
- **Codex CLI** — reference via `-p` flag
## How to Contribute
1. Fork the repo
2. Add your skill as `skills/your-skill-name/SKILL.md`
3. Open a PR with description of the DevOps pain point it solves
## 🧬 Skill Anatomy
Each skill lives in its own directory under `skills/` and follows a consistent structure:
```
skills/
├── self-regulation-brake-system/
│ └── SKILL.md
├── macos-ram-swap-analysis/
│ └── SKILL.md
├── github-exposure-standard/
│ └── SKILL.md
└── ... (14 total)
```
Every `SKILL.md` is composed of two parts:
**YAML Frontmatter** — machine-readable metadata that agents use for discovery and matching:
```yaml
---
name: macos-ram-swap-analysis
description: Diagnose memory pressure and swap congestion on macOS
tags: [macos, memory, swap, diagnostics]
triggers:
- memory full
- swap usage
- ram pressure
---
```
**Markdown Body** — the human-readable playbook the agent follows, typically structured as:
- `## Problem` — what pain point this skill solves
- `## Solution` — the exact commands, scripts, or procedures to run
- `## Integration Points` — how to wire it into the agent's workflow (config files, hooks, env vars)
This dual format means agents can **auto-discover** skills via frontmatter tags/triggers while humans can **read and understand** the full procedure. No lock-in, no opaque configs.
## License
MIT — use freely, improve openly.