Pricing
Every plan runs all 124 checks. Upgrade to Pro for the corrected code, step-by-step fixes, A2SPA guidance, and a PDF report you can hand to clients and auditors.
Secure checkout with Stripe · Cancel anytime · No credit card for Free
Try every check on your agent and see exactly what is wrong.
10 scans/month
Free does not include
Fix what the scan finds — with the corrected code — and prove it with a report.
100 scans/month
For teams shipping several agents who want one shared place to review them.
2-seat minimum
For regulated teams and private agent security programs.
How scans are counted
Free includes 10 scans/month; Pro includes 100. Dashboard, CLI, API, and GitHub Action scans all share the same monthly quota, and scanning a whole GitHub repo or .zip counts as one scan.
An OWASP Agentic Skills assessment uses one scan from the same monthly quota, whether or not you ask for a signature.
Compare plans
Every plan runs the same 124 checks — we never hide a risk behind a paywall. Pro is for when you need to actually fix what was found, quickly, and show proof that you did.
| Feature | Free | Pro · $19.99 | Team · $25/seat |
|---|---|---|---|
| Find the problems | |||
| Scans per month | 10 | 100 | 300 per member |
| GitHub PR bot + monitoring: scans per month | 50 | 1,000 | 1,000 |
| All 124 security checks (incl. framework + MCP packs) | ✓ | ✓ | ✓ |
| Verdict, 0–100 security score, prioritized findings | ✓ | ✓ | ✓ |
| Scan a GitHub repo, a .zip, many files, or an MCP config | ✓ | ✓ | ✓ |
| OWASP Agentic Skills assessment (+ optional signature) | ✓ | ✓ | ✓ |
| Compliance mapping: OWASP LLM Top 10, NIST AI RMF, SOC 2 | ✓ | ✓ | ✓ |
| Fix them fast | |||
| Step-by-step fix guidance for every finding | — | ✓ | ✓ |
| Corrected code you can copy in | — | ✓ | ✓ |
| Fix & re-check: AI fixes your file, re-check instantly without using a scan | — | ✓ | ✓ |
| A2SPA execution-authorization guidance | — | ✓ | ✓ |
| Copyable fix checklist (GitHub / Jira / Slack) | ✓ | ✓ | ✓ |
| Ship with proof | |||
| PDF security report for clients, auditors, and leadership | — | ✓ | ✓ |
| Shareable report link + verified badge | ✓ | ✓ | ✓ |
| CLI + GitHub Action to block risky merges | ✓ | ✓ | ✓ |
| VS Code extension + MCP server for Claude Code / Cursor | ✓ | ✓ | ✓ |
| GitLab, Bitbucket and Slack | ✓ | ✓ | ✓ |
| Agent Guard runtime protection (allow lists, approvals, dashboard) | ✓ | ✓ | ✓ |
| One-click fix suggestions in pull requests | — | ✓ | ✓ |
| Prompt-injection tests against your own agent | — | ✓ | ✓ |
| Compliance evidence pack (EU AI Act, ISO 42001, SOC 2, NIST) | — | ✓ | ✓ |
| Fixes inside PR comments and VS Code quick fixes | — | ✓ | ✓ |
| Repo monitoring: score history + Slack alerts on new risks | ✓ | ✓ | ✓ |
| Work as a team | |||
| Shared workspace with roles | ✓ | ✓ | ✓ |
| Pro features for every member | — | — | ✓ |
| Onboarding call + priority support | — | — | ✓ |
What Pro unlocks
Free
Critical · SQL query built from unescaped input
A SQL statement is assembled by string concatenation instead of bound parameters. Line 14.
db.query('SELECT * FROM invoices WHERE id = ' + invoiceId)You know what's wrong — now you research the fix yourself.
Pro
Critical · SQL query built from unescaped input
Use bound parameters for every value, and give the agent a read-only database role limited to the tables it needs.
db.query('SELECT * FROM invoices WHERE id = $1', [invoiceId])Copy the fix, re-scan to confirm, export the PDF for your records.
Minutes, not hours
Corrected code for each finding instead of researching every fix from scratch.
Scan every change
100 scans a month covers re-scanning after each fix and every pull request in CI.
Proof for buyers
A PDF security report your clients, auditors, and leadership can read.
Yes. New Pro customers get 14 days free. Checkout asks for a card, and you are only charged when the trial ends — cancel any time before then in Manage billing and you pay nothing.
No. Every plan runs all 124 checks. Free shows you what is wrong; Pro adds the step-by-step fix, corrected code, A2SPA guidance, and PDF export.
Yes, on any plan. Create an API key in the dashboard, then run the CLI or the GitHub Action — risky changes fail the check before they merge. See the Developers page.
Monthly scans (10 Free, 100 Pro, 300 per Team member): each scan you run from the dashboard, the CLI, the API or the GitHub Action, and each OWASP assessment. Scanning a whole repo or .zip is still one scan. Pull-request checks and repo monitoring from the GitHub, GitLab and Bitbucket integrations use a separate allowance (50 a month on Free, 1,000 on Pro and Team), and Slack scans have their own (20 a month on Free, 1,000 on Pro). Re-checks in Fix & re-check and the public demo do not count.
Scans run on Agent Verify’s server, never in your browser. Your raw code is not stored and is never used to train AI models; we keep the results (verdict, score, findings) so your history works, with secret values masked. Reports stay private until you share them. Details are on the Trust page.
You pick the number of seats (from 2). Everyone you add gets every Pro feature and 300 scans a month, and you can add or remove people yourself from Billing.
Install the Agent Verify GitHub App and every pull request that changes agent code gets a pass/fail check, notes on the risky lines, and one comment that updates on each push. Pro adds the fixes to that comment.
Yes. Manage or cancel Pro from the billing portal; you keep Pro until the end of the period you paid for.