Documentation preview. These pages target Registryctl v0.15.0, which is not published yet. Public download commands will not work until that release exists. For runnable released instructions, use v0.13.0.
Use Registry Stack when an institution already manages authoritative data and needs a controlled way for another system to read selected records or receive a bounded answer. Registry Stack adds an access layer. It does not replace the system that owns the data.
Registry Stack is a good fit when
Section titled “Registry Stack is a good fit when”- The source must remain authoritative and under the institution’s control.
- Callers need a reviewed interface instead of direct database or file access.
- Each caller must have an explicit purpose, permission, and limited result.
- The institution needs to test source behavior with synthetic fixtures before connecting a runtime environment.
- Operators need separate control over endpoints, credentials, trust, and deployment.
Choose the result you need
Section titled “Choose the result you need”| Caller needs | Use | Result |
|---|---|---|
| Selected records or fields | Registry Relay | A protected, read-only API response |
| A bounded answer or status | Registry Notary | A claim result without the source record |
The two products can work together. Registry Relay obtains a limited source result, and Registry Notary evaluates a reviewed claim over that result.
Registry Stack is not the right tool when
Section titled “Registry Stack is not the right tool when”- You need a new system of record.
- A caller must create, update, or delete source records.
- You need a workflow, eligibility, ranking, or case-management engine.
- You need Registry Stack to decide legal authority, data ownership, or institutional approval.
- You want to publish unrestricted open data.
Registry Stack makes these boundaries visible, but the responsible institutions still make and approve the underlying decisions.
What you need to adopt it
Section titled “What you need to adopt it”Before moving beyond a local sample, identify:
- The source owner and update process.
- The fields and operations the integration is allowed to use.
- The callers, purposes, and fields each service may disclose.
- The operator responsible for endpoints, credentials, trust, deployment, monitoring, recovery, and audit retention.
Start with non-personal data. A successful local tutorial is not production approval.