The #1 security scanner for AI agents and LLM applications. Detect API key leaks, prompt injections, CVE vulnerabilities, and misconfigurations in seconds.
From API key leak detection to prompt injection scanning, Keprax provides comprehensive LLM security scanning for your AI agents.
35+ API key formats: OpenAI, Anthropic, AWS, Stripe, GitHub, and more.
Checks dependencies against OSV database for known vulnerabilities.
Catches jailbreaks, instruction overrides, and encoding bypasses.
Track your security score over time with visual trends.
Earn XP, unlock achievements, climb the leaderboard.
Embeddable badge for your README to show security score.
Auto-scan on every push with CI/CD integration.
Run `keprax scan .` from your terminal with pre-commit hooks.
Daily CVE digest tailored to your tech stack.
Collaborate across your org with shared scans and history.
Create fake API keys that alert you when someone uses them.
Get notified on critical findings in your team channels.
Detects webhook leaks, DNS tunneling, and suspicious patterns.
"What if this leaked?" - Impact analysis for exposed secrets.
Auto-generate HackerOne/Bugcrowd formatted reports.
Start free. Upgrade when you need more.
Paste code, upload files, connect GitHub, or use the CLI.
Keprax analyzes for secrets, CVEs, injections, and misconfigs instantly.
Get remediation steps, track progress, earn achievements.
Works with every AI agent platform
Everything you need to know about AI agent security scanning
One exposed API key can cost thousands. Start AI agent security scanning for free.