Residency & Training
Check what is publicly documented about data region and training use.
Agent Check · concrete deployment review
From requirements to an evidence-backed shortlist.
Agent Check connects concrete must-have requirements with public agent and governance evidence. It shows full matches, evidence gaps and explicit conflicts.
Direct answer
Agent Check evaluates requirements against evidence — not marketing claims.
A match means suitable public evidence exists for every selected criterion within the documented scope. Unknown remains an evidence gap. A documented “No” is only a conflict when your requirement actually expects “Yes”; for “customer data for training”, “No” may explicitly be the desired result.
Agent Check · EU & Governance
Check must-have requirements without treating Unknown as No.
The check uses the same field-level evidence as the trust panels on agent profiles. Privacy, data residency, access control, human oversight and auditability are especially visible.
36Trust Controls
4evidence pillars
0blanket scores
1:1Signal → source
RBAC, SSO, encryption
Check documented access and protection mechanisms.
Human oversight
Check Human Approval and other control mechanisms.
Audit & Traceability
Check logs, SIEM, APIs and traceability.
Direct answer
The check is a pre-assessment: which agents have documented evidence for your must-have requirements, and where do you need deeper review?
No paid rankingsUnknown ≠ falseScope stays visibleprimary sources over product claims
Step 1
Choose your must-have requirements
Multiple criteria are checked together. A full match needs suitable evidence for every selected criterion.
Requirement in natural language
Describe what your deployment absolutely requires.
Example: “EU Data Residency, Human Approval, no customer data for training, Audit Logs and SSO.” The parser only maps known terms to structured controls; it does not invent requirements.
No text requirement evaluated yet.
Step 2
Read the results
Select at least one requirement. Unknown is shown as an evidence gap, not as No.
Evidence instead of a score
The check uses the same trust signals as the agent profiles.
Each selected must-have requirement is checked against field-level governance evidence for the agent. A match shows scope, review date and primary source. If robust evidence is missing, the state remains Unknown. This prevents the check from turning missing documentation into a supposed negative property.
1 · Must-have requirement
You select criteria such as Data Residency, RBAC, Human Approval or Audit Logs.
2 · Evidence matching
AgentenCode searches only explicit field-level signals within the documented scope.
3 · Return to profile
The result links back to the agent profile, where the same trust signals and sources are visible.
Interpretation
Three different result types.
They must not be mixed together.
Requirement documented
Suitable evidence exists for the selected criterion within the stated scope.
Not sufficiently documented publicly
The agent may still have the property; the current dataset simply does not support the claim.
Explicitly documented against the requirement
A documented conflicting value is stronger than Unknown and should be reviewed deliberately.
Review chain
What comes after Agent Check.
A robust approval depends on the concrete system state, not just the product name.
Shortlist
Narrow agents by documented must-have requirements.
Check scope
Compare plan, region, deployment, workspace and provider conditions.
Check system state
Review identity, permissions, delegation, approval, audit, rollback, memory and hosting.
Internal approval
Perform internal security, privacy, legal and domain review.
No hidden product rating
Agent Check does not calculate a universal quality or compliance score. An agent with less public documentation may still fit a concrete use case, while a well-documented agent may still fail a project-specific must-have requirement.
Is Agent Check a purchase recommendation?
No. The check shows coverage of your selected requirements in the documented evidence base.
What does evidence gap mean?
There is not enough public evidence for the selected criterion within the examined scope.
Does the check replace a compliance review?
No. Contractual, legal, privacy and security review remain dependent on the concrete deployment context.
EU / GDPR context
Agent Check evaluates evidence requirements, not legal compliance. For EU AI Act and GDPR context, see the AgentenTrust EU layer.