Skip to main content

AI testing & review

Guardana

Open-source AI security verification

Visit guardana.dev (opens in a new tab)0 votes. Log in to vote for Guardana in print
  • Community votes0
  • DR by Ahrefs2 no change since last check

Ahrefs Domain Rating set in type 25 Oct 2026, reset 25 Oct 2026.

Guardana metrics

Set in type for guardana.dev: Ahrefs Domain Rating and Majestic link data, with the change since the last edition.
MetricValue
Domain Rating2 no change since last check
Trust Flow2 no change since last check
Citation Flow21 no change since last check
Referring domains22 no change since last check
External backlinks271 no change since last check

Ahrefs Domain Rating set in type 25 Oct 2026, reset 25 Oct 2026. Lead topic: Computers/Software/Operating Systems (2). Majestic set 25 Oct 2026, reset 25 Oct 2026. Link data typeset by Majestic. Trust Flow, Citation Flow and Topical Trust Flow are trademarks of Majestic-12 Ltd.

Guardana domain colophon

From the registry's public RDAP record, checked 25 Oct 2026, 14:51 UTC.
Domain age3 months
Registered
Expires
RegistrarGoDaddy.com, LLC
DNSSECNot signed

Full colophon for guardana.dev

Tool for security and platform engineers to verify AI systems by scanning model artifacts, testing live endpoints, and grading recorded runs.

The story so far

Screenshot of the Guardana homepage
Homepage of Guardana

Guardana is an open-source AI security verification tool that helps security and platform engineers check model and code artifacts, live AI systems, and recorded runs. It provides four main capabilities: scan (check artifacts offline), probe and monitor (test live endpoints), grade (evaluate recorded answers), and diff (compare security runs). The tool includes 58 rules across build-time and runtime checks, runs without requiring an account or telemetry, and can be integrated into CI pipelines or called from Python code. A companion project, Guardana Control, provides runtime MCP tool-call enforcement.

Typeset features

  • Scan model and code artifacts offline
  • Probe and monitor live endpoints
  • Grade recorded application runs
  • Diff and compare security runs
  • 58 security rules across build and runtime
  • Python API for integration
  • Custom rule creation with YAML or Python
  • Cost estimation before probing
  • Optional self-hosted collector

Pricing

Pricing model: Open source

Compare and explore

Questions and answers

Does a scan need internet access?

No. Artifact scans run fully offline. Results leave the machine only when --reporter names a collector or reporter.

Which models does Guardana work with?

Guardana works with any model format for scans. For probes, it addresses served endpoints regardless of model, including OpenAI-compatible APIs, Ollama, Hugging Face TGI, custom APIs, MCP servers or A2A agents.

Can I grade the same replies with a new rule or judge?

Yes. Run guardana grade RECORDING again with a sharper rule or another judge without calling the application again.

Can I call Guardana from code or over HTTP?

Yes. guardana.core.verify runs a scan, probe or grade from Python and returns data. The optional collector's HTTP API stores and serves results but does not start scans or probes.

How does Guardana relate to Guardana Control?

They are independent projects. Guardana scans and probes outside the request path and grades runs in its own trace format. Guardana Control is an alpha MCP tool-call gate that enforces verdicts and records evidence. Neither needs the other.

Find Guardana online

Leave a review for Guardana

No review of Guardana set in type yet. Set the first.

Your rating

Reviews are free to set. None are paid for. Read the proofs

Readers' notes · Reprints · Share the page · Your company in print? Claim it · Report a misprint

More on this profile

Guardana has a free listing. The owner can unlock the full profile with a paid launch for $5:

  • Founder story (locked)
  • Milestones (locked)
  • Demo video (locked)
  • Dofollow website link (locked)

More pages like this

  1. Number 1: GPTZero logoGPTZero

    AI detector that identifies content generated by large language models

    AI testing & review

    DR 81 no change since last check 0 votes. Log in to vote for GPTZero in print

  2. Number 2: Originality.ai™ logoOriginality.ai™

    Most accurate AI detector for identifying ChatGPT, GPT-5, Gemini and other AI-generated content

    AI testing & reviewCanada

    DR 79 no change since last check 0 votes. Log in to vote for Originality.ai™ in print

  3. Number 3: 99tests

    Real testers run your AI-generated test cases on real phones and browsers with a human-verified pass/fail report in 3 days

    AI testing & reviewIndia

    DR 27 no change since last check 0 votes. Log in to vote for 99tests in print

  4. Number 4: Papercuts logoPapercuts

    Test in Production

    AI testing & review

    DR 1 no change since last check 0 votes. Log in to vote for Papercuts in print
Help centre