Wytness — The Blackbox Recorder for AI Agents
wytness · Intelligence & Research
Certification per Microsoft Marketplace.
Evidence tier Source Confirmed · 1 capture on record
What the publisher says
As described on Microsoft Marketplace.
Wytness is the managed compliance and audit layer for AI agents. Every action
your agents take is signed with your key, every sensitive payload is sealed so we never see it, and
Show the rest of the publisher’s description (30 more lines)
the whole record is auditor-ready. Built on Microsoft's open-source Agent Governance Toolkit (AGT),
Wytness is vendor-neutral and runs wherever your agents do.
Why teams running regulated AI agents choose Wytness:
- You hold the keys. Personally identifiable data is pseudonymised in your own
environment with a key only you hold. Wytness stores evidence without ever seeing the secrets —
true zero-knowledge.
- Non-repudiable evidence. Every event carries a per-event Ed25519 signature in
a tamper-evident envelope. An agent owner attested to this exact action and cannot deny it later.
- Auditor-ready output. Evidence Packs map your agent activity to EU AI Act,
SOC 2, and ISO 27001 controls — the deliverable your auditor actually expects.
- Your storage, or ours. On Business and above, bring your own storage —
evidence lands in an Azure Blob container in your own tenancy, where you can apply your own
immutability policy. AWS S3 is on the roadmap. Or let Wytness host it for you.
- Retention that matches your obligations. A live queryable dashboard window on
every plan, and from Business upward a seven-year archive with signed handover on exit.
- Choose where your evidence lives. Data residency is available on Business and
above — tell us the region you need and we provision into it.
Three surfaces, one platform:
- Ledger — the audit and evidence layer: per-event signing, customer-held PII
keys, long-term retention, and Evidence Pack PDFs.
- Registry — the inventory and detection layer: a live dashboard of Events,
Agents, Tools, Sessions and Permissions, with a seven-rule behavioural anomaly engine.
- Shield — the policy layer: author policies and run them in observe mode, with
in-process enforcement available through the AGT library in your own runtime. Wytness-side enforce
mode and human-in-the-loop approval routing arrive Q4 2026.
Built on Microsoft AGT. Wytness extends Microsoft's Agent Governance Toolkit —
identity, delegation and a tamper-evident hash chain — with the non-repudiation and privacy layer
regulated buyers need. Install the Wytness SDK (Python or TypeScript), and your existing AGT audit
events flow straight into Wytness, signed and sealed.
Fully managed SaaS. Up and running in minutes. Cloud-hosted and regulator-ready.
Preview
5 images
Agent build and provenance
See the full provenance
The layer-by-layer build, the evidence behind each claim, the risk basis and the cross-marketplace links are open to any account. Some rows are disclosed, some the source leaves Unknown; a free account shows you which.
Plans and pricing as listed
3 listed- Active Agents: $39.00 per agent / month
- Audit Events: $0.002 per audit event
- Active Agents: $39.00 per agent / month
- Audit Events: $0.002 per audit event
- Active Agents: $29.00 per agent / month
- Audit Events: $0.001 per audit event
- Active Agents: $29.00 per agent / month
- Audit Events: $0.001 per audit event
- Active Agents: $19.00 per agent / month
- Audit Events: $0.0005 per audit event
- Active Agents: $19.00 per agent / month
- Audit Events: $0.0005 per audit event
Sources
Publisher resources
2 linksLinked repositories
Unknown means this listing does not publish a repository. It is not a statement that the code is closed, and a linked repository is not a claim that the publisher wrote it: the registry computes that relationship privately and does not publish it.
Evidence risk is the share of the build you cannot see before you deploy, not a security rating. Sign in to see the layer-by-layer basis for this band.