Config Guard — AI skill for Claude Code

Config Guard 🛡️

[🇨🇳 中文文档](./README_CN.md)

[](https://clawdhub.com) [](./LICENSE) [](./SKILL.md)

Stop AI agents from killing themselves by editing their own config wrong.

Prompts are suggestions. Code is law.

OpenClaw agents have full access to `openclaw.json` — the file that controls the gateway, channels, models, and tools. One bad edit = gateway crash. Agent goes offline. Nobody can fix it remotely.

This skill prevents that.

The Problem

Real incidents (all from the same day):

What happened	Root cause
Gateway crashed after update	AI added unknown config fields (auth, fallbacks)
Model not found	AI wrote claude-sonnet-4.5 (dots) instead of claude-sonnet-4-5 (hyphens)
Telegram stopped working	Config change silently wiped the channel section
browser.profiles error	AI forgot the required color field (must be hex)
Plugin SDK missing	Update broke module paths, no validation caught it

**7 cascading failures. 1 root cause: AI doesn't know correct config formats. It guesses.**

What It Does

Before config changes:

1. Auto-backup — timestamped copy before any change
2. JSON syntax check — catches malformed JSON
3. Schema validation — required fields, correct types, valid values
4. Semantic checks — catches AI-specific mistakes:
   • Model names with dots instead of hyphens
   • Missing required fields (browser.profiles.color)
   • Unknown top-level keys
   • Placeholder values in sensitive fields
5. Critical field diff — warns if Telegram channel, auth tokens, or tool deny lists changed

After config changes:

1. Gateway health check — polls for up to 30s
2. Auto-rollback — if gateway doesn't recover, restores backup automatically

Quick Start

# Instal