Openclaw Ops Skills

# OpenClaw Ops Skills Pack

> **Production-ready skills for autonomous agent operations**

## Metadata

**Name**: Ops Skills Pack
**Version**: 1.0.0
**Category**: Operations
**Author**: Eric Jie <jxmerich@mail.com>
**License**: MIT
**Compatibility**: OpenClaw >= 2.0.0
**Repository**: https://github.com/Erich1566/openclaw-ops-skills

## Description

Transform your OpenClaw agent from a chatbot into production-ready autonomous infrastructure. Based on 1.4+ billion tokens of real-world testing, this skill pack provides the guardrails, workflows, and best practices needed for autonomous operations that work while you sleep.

## Skills Included (14)

### Core Infrastructure
1. **model-routing** - Optimize model selection for 80% cost reduction
2. **execution-discipline** - Enforce Build→Test→Document→Decide cycle
3. **docs-first** - Mandatory reconnaissance before code changes
4. **scope-control** - Define and enforce operational boundaries
5. **task-autonomy** - Self-expanding task management
6. **progress-logging** - Comprehensive execution logs

### Operations
7. **memory-persistence** - File-based memory system
8. **cron-orchestration** - Scheduled autonomous operations
9. **error-recovery** - Graceful failure handling
10. **security-hardening** - Security best practices

### Advanced
11. **cost-optimization** - Token waste prevention
12. **testing-protocol** - Quality assurance standards
13. **communication** - Structured update templates
14. **integration-guide** - Safe third-party connections

## Key Benefits

- **80% Cost Reduction** - Intelligent model routing (Sonnet 4.6 for daily work)
- **True Autonomy** - Self-expanding task lists continue work overnight
- **Persistent Memory** - Survives session compression
- **Complete Visibility** - Morning briefings replace chat history scrolling
- **Security Hardened** - Built-in security practices

## Quick Start

```bash
# Install skills
cp skills/*.md ~/.openclaw/wo...

# OpenClaw Ops Skills Pack

> **Production-Ready Skills for OpenClaw Infrastructure**
> Version: 1.0.0 | Last Updated: 2026-03-10

## Overview

This is a comprehensive skill package designed to transform OpenClaw from a chatbot into reliable infrastructure. Based on 1.4 billion tokens of real-world testing, these skills provide the guardrails, workflows, and best practices needed for autonomous agent operations.

## What This Solves

| Problem | Solution |
|---------|----------|
| Agent loops on same answer 8 times | `execution-discipline.md` |
| Agent edits files it shouldn't touch | `scope-control.md` |
| Agent skips documentation and breaks things | `docs-first.md` |
| Agent stops after one task and waits | `task-autonomy.md` |
| Lost context when sessions compress | `memory-persistence.md` |
| Massive token waste on wrong models | `model-routing.md` |
| Security vulnerabilities | `security-hardening.md` |
| No visibility into overnight work | `progress-logging.md` |

## Quick Start

```bash
# Clone to your OpenClaw workspace
git clone https://github.com/yourusername/openclaw-ops-skills.git ~/.openclaw/workspace/skills/ops-pack

# Or copy individual skills
cp skills/*.md ~/.openclaw/workspace/skills/

# Restart OpenClaw
openclaw restart
```

## Skills Included

### Core Infrastructure Skills

| Skill | Purpose | Priority |
|-------|---------|----------|
| `model-routing.md` | Optimize model selection for cost/quality | ⭐⭐⭐ |
| `execution-discipline.md` | Build → Test → Document → Decide loop | ⭐⭐⭐ |
| `docs-first.md` | Force documentation before code changes | ⭐⭐⭐ |
| `scope-control.md` | Define and respect operational boundaries | ⭐⭐⭐ |
| `task-autonomy.md` | Self-expanding task management | ⭐⭐⭐ |
| `progress-logging.md` | Comprehensive progress tracking | ⭐⭐⭐ |

### Operational Skills

| Skill | Purpose | Priority |
|-------|---------|----------|
| `memory-persistence.md` | Maintain state acr...