Free safety scores for every AI product, agent and service.
Know what's safe before you use it.
There's no universal safety rating for AI tools. You're left to trust the marketing, read Reddit threads, or just hope for the best.
Many AI tools request access to your files, contacts, and personal data. Most people click "allow" without understanding what they're giving away.
A slick landing page doesn't mean a tool is safe. Malicious AI services look identical to trustworthy ones. You can't tell the difference at a glance.
A tool that was safe last month can push a bad update, change its privacy policy, or start behaving differently — and nobody tells you.
// Check if a tool is safe before granting access const res = await fetch( 'https://api.fabriclayer.ai/v1/lookup?slug=mcp-python-sdk' ) const { score, status, signals } = await res.json() // score: 4.59 // status: "trusted" // signals: // vulnerability: 4.8 // maintenance: 5.0 // operational_health: 4.5 // adoption: 3.9 // transparency: 4.5 // publisher_trust: 4.3 if (status === 'blocked') { throw new Error('Tool failed safety check') }