Agent Passport Pro

Real-time threat definitions for AI agents.

Like antivirus signature updates, but for your AI agent stack. 75+ security patterns across skill scanning, injection detection, SSRF protection, and path traversal defense. Auto-updated every 6 hours. No manual intervention.

Coming Soon

We're putting the finishing touches on our production environment.
Join the priority waitlist for early access.

We'll email you when Pro is ready · legal@agentpassportai.com for questions


What you get

Feature Free Pro
Mandate system, KYA, Kill Switch, Audit Yes Yes
scan-skill (pre-install scanner) Yes Yes
check-injection (runtime protection) Yes Yes
Threat definitions (scan + injection + SSRF + path traversal) Static (baked in at release) Live: auto-updated every 6 hours
Pattern coverage 28 scan + 20 injection patterns 75+ patterns, growing weekly
Weekly threat digest email Every Monday
Approval relay DIY (self-hosted Telegram bot) Hosted, zero setup
Audit dashboard 7-day local JSON 90-day web UI
Support Community Email (48hr)

Setup takes 30 seconds

After purchase, you'll receive a license key by email. Add it to your environment:

# Add to ~/.bashrc or ~/.zshrc
export AGENT_PASSPORT_LICENSE_KEY="ap_pro_your_key_here"

# That's it. Definitions auto-update in the background:
mandate-ledger.sh scan-skill ./any-skill/
# Threat definitions: 2026.02.26.1 (pro-auto)

# Or check your status anytime:
mandate-ledger.sh definitions-status

Definitions auto-update silently in the background every 6 hours. Your scans are never delayed. The license key is cached locally for 7 days, so offline usage works fine. If you cancel, the skill falls back to static patterns at the end of your billing period.


Common questions

What happens if I cancel?

Your key stops validating at the end of the billing period. The skill falls back to static patterns automatically. Nothing breaks. No data is deleted.

Does Agent Passport phone home?

Pro makes two types of network calls: license validation (sends your key, returns your tier) and threat definition updates (pulls latest patterns). Both happen at most every 6 hours and are cached locally. No usage data, no telemetry, no scan results leave your machine. Free tier makes zero network calls.

Can I use one key on multiple machines?

Yes. One Pro license covers all machines you use personally or with a small team (up to 5 agents). Contact us if you need more.

What are live threat definitions?

Agent Passport ships with 75+ security patterns covering skill scanning, prompt injection detection, SSRF protection, and path traversal defense. Free users get these patterns frozen at each release. Pro users get new definitions automatically every 6 hours. When a new malicious skill is discovered on ClawHub (Snyk found 76 confirmed payloads in their February 2026 audit), Pro users are protected within hours. Free users wait for the next release.

How does the auto-update work?

Every time your agent runs a scan, Agent Passport checks if your local definitions are more than 6 hours old. If they are, it pulls fresh definitions from our API in the background. Your scan completes immediately using cached definitions. The update happens silently. No restarts, no manual steps.

Is the code still open source?

Yes. The local skill code is MIT with Commons Clause: free to use and modify, not to sell. The hosted API services are proprietary. See LICENSE and TRADEMARK.md on GitHub.