princezuda/safeclaw
Safeclaw is the alternative to openclaw.. You can naturally chat with it via text and voice, and you can choose not to use a language model., By default it picks up on intent and semantics.. No prompt injection while you get over ninety percent of what openclaw does plus tts and voice to text
Built on traditional ML pipelines (VADER sentiment, spaCy NLP, YOLO object detection, Whisper/Piper for local speech), SafeClaw achieves 90% of OpenClaw's capabilities without LLM dependencies—meaning zero API costs and minimal prompt injection surface. Integrates across email (IMAP/SMTP), calendars (ICS/CalDAV), smart home devices, Bluetooth, RSS feeds, and social platforms (Twitter, Mastodon, Bluesky), with multi-channel deployment via CLI, Telegram, and webhooks; optional LLM routing (Anthropic, Ollama) available only for research and blogging tasks.
105 stars.
Stars
105
Forks
10
Language
Python
License
MIT
Category
Last pushed
Mar 12, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/agents/princezuda/safeclaw"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
LeoYeAI/openclaw-guardian
🛡️ Guardian watchdog for OpenClaw Gateway — auto-monitor, self-repair via doctor --fix,...
Ramsbaby/openclaw-self-healing
AI-powered self-healing system for OpenClaw Gateway • 4-tier autonomous recovery • macOS & Linux
jzOcb/config-guard
Stop AI agents from crashing OpenClaw by validating config changes. Auto-backup, schema...
grabee-chen/openclaw-security-starter
🦞 Security-first starter template for OpenClaw agents — 4-layer protection against prompt...
Zjianru/restart-guard
OpenClaw Skill: Safely restart the Gateway with context preservation, guardian watchdog, and...