Scan. Fix. Harden. Get a security score out of 100 for your AI agent installation. Detect CVEs, misconfigurations, and exposed secrets in seconds — powered by Bankr LLM Gateway.
A comprehensive security toolkit built for the AI agent ecosystem.
Cross-references your agent version against known CVEs with CVSS scores, affected version ranges, and patch guidance.
One command to harden your installation. Fixes permissions, generates strong tokens, patches config, and removes dangerous flags.
Scans for hardcoded API keys, tokens, and passwords across 10+ secret patterns including OpenAI, GitHub, AWS, Slack, and more.
Leverages 27+ LLMs via Bankr LLM Gateway for human-readable explanations and prioritized fix recommendations.
Built in Rust. Ships as one self-contained binary — no runtime, no dependencies, no Docker required.
Validates 30+ configuration settings against security best practices — auth, sandboxing, network exposure, and more.
Install, scan, and fix. It takes less than a minute.
Every check targets a real attack surface. No filler, no vanity metrics.
| Category | Checks | Points |
|---|---|---|
| Authentication | 3 | 12 |
| Known Vulnerabilities (CVE) | 15 | 51 |
| Secrets Management | 5 | 12 |
| File Permissions | 3 | 10 |
| Network Exposure | 3 | 12 |
| Tool Authorization | 3 | 10 |
| Exec Security | 3 | 10 |
| Sandboxing | 3 | 10 |
| DM Security | 2 | 8 |
| Group Security | 2 | 6 |
| Browser Security | 2 | 6 |
| Dangerous Flags | 1 | 6 |
| Logging & Privacy | 2 | 4 |
| mDNS / Discovery | 1 | 2 |
| Control UI | 2 | 2 |
| Plugins | 1 | 2 |
| Total | 54 | 163 |
Precompiled binary for Linux x86_64. Works on WSL, Ubuntu, Debian, and most Linux distros.
Star us on GitHub and help secure the AI agent ecosystem.