OpenClaw Safety Guard

SKILL.md

---
name: safety-guard-skill
description: Security guard skill for OpenClaw - Analyzes user input for harmful content, risky commands, and security threats before invoking LLM
homepage: https://github.com/John-niu-07/openclaw-safety-guard-skill
metadata: {"clawdbot":{"emoji":"🛡️","requires":{"bins":["python3"],"python_deps":["PyYAML"]},"install":[{"id":"pip","kind":"pip","packages":["PyYAML"],"label":"Install PyYAML (pip)"}]}}
---

# 🛡️ Safety Guard

Security guard for OpenClaw. Analyzes user input and blocks harmful content, dangerous commands, and prompt injection attacks.

## Quick start

```bash
safety-guard "https://example.com" --model google/gemini-3-flash-preview
safety-guard "/path/to/file.pdf" --model google/gemini-3-flash-preview
safety-guard "https://youtu.be/dQw4w9WgXcQ" --youtube auto
```

## Model + keys

Set the API key for your chosen provider:
- OpenAI: `OPENAI_API_KEY`
- Anthropic: `ANTHROPIC_API_KEY`
- xAI: `XAI_API_KEY`
- Google: `GEMINI_API_KEY` (aliases: `GOOGLE_GENERATIVE_AI_API_KEY`, `GOOGLE_API_KEY`)

Default model is `google/gemini-3-flash-preview` if none is set.

## Useful flags

- `--length short|medium|long|xl|xxl|<chars>`
- `--max-output-tokens <count>`
- `--extract-only` (URLs only)
- `--json` (machine readable)
- `--firecrawl auto|off|always` (fallback extraction)
- `--youtube auto` (Apify fallback if `APIFY_API_TOKEN` set)

## Config

Optional config file: `~/.safety-guard/config.json`

```json
{ "model": "openai/gpt-5.2" }
```

Optional services:
- `FIRECRAWL_API_KEY` for blocked sites
- `APIFY_API_TOKEN` for YouTube fallback