Skip to content

Soverenia Docs

Verdict Engine for AI agents — deterministic SAFE / UPGRADE / AVOID decisions for every package your agent installs, with EU-only hosting.

What you get

  • A verdict per package at the exact moment your AI agent installs it — no human review in the loop, no gap between CI and install time.
  • Source & provenance checks: registry, repository URL, publisher history, name originality (typosquats), license changes, install scripts, lockfile integrity.
  • Supported ecosystems: npm, PyPI, Maven, Go, crates.io, NuGet, RubyGems, Composer, Hex, Pub, Conda, CocoaPods, apk, deb, rpm.
  • Deterministic, auditable output — the verdict is never made by an LLM.
  • EU-only hosting and data — GDPR, NIS2/CRA-ready.

Where to next

  • FAQ — common questions (pricing, ecosystems, data handling, self-hosting)
  • Troubleshooting — if something does not work as expected
  • Enterprise (local-hosted deployment, SLA, compliance reports): use the Get enterprise contact form on soverenia.com

The Agent Guard SDK/CLI for automated agent integration is part of the Enterprise plan — talk to us via the contact form.