Integration Ownership Model
Use this page when you need to know which repo or runtime owns an integration behavior before you implement or troubleshoot it.
When to use this page
- You are deciding where a change belongs.
- You need to avoid fixing an integration symptom in the wrong repo.
- You need a public-safe handoff model for Helpifyr-connected systems.
Prerequisites
- You have read System Model.
- You know which product or runtime is involved.
Architecture / Flow
flowchart LR
A["Integration question"] --> B["Find canonical owner repo"]
B --> C["Find Fabric shared truth surface"]
C --> D["Find docs or runtime consumer"]
D --> E["Apply change in owner lane"]
Step-by-step procedure
1. Start with the owner question
Ask:
- Which repo defines the contract?
- Which runtime exposes the live behavior?
- Which docs repo only materializes or publishes the reviewed result?
2. Apply the standard ownership split
helpifyr-fabric- owns shared control-plane routes, contracts, projections, and public docs-platform truth
jhf-docs- owns reviewed public-docs materialization and repo-owned safe overrides
jhf-web- owns the live site publishing and deployment of the reviewed artifact
- downstream product repos
- own product-specific runtime behavior and local contracts that Fabric consumes or summarizes
3. Use Fabric as the shared readback layer
Illustrative routes:
GET /api/v1/contracts/registry
GET /api/v1/platform/projection-catalog
GET /api/v1/tools/runtime-status
GET /api/v1/docs/platform
These routes tell you what Fabric currently believes and publishes, but they do not automatically make Fabric the mutation owner for every downstream behavior.
4. Send the fix to the true owner lane
Examples:
- stale public page structure:
- likely
jhf-docsorjhf-web
- likely
- missing or wrong shared route documentation:
- likely
helpifyr-fabric
- likely
- product-specific runtime mismatch:
- likely the downstream owner repo or environment
Verification
Ownership is classified correctly when:
- the canonical repo is explicit
- the readback route is explicit
- the next mutation does not happen in a non-owner layer
Common failure modes
Fixing a shared-truth problem only in the published site
Problem:
- the symptom disappears briefly but source truth stays wrong.
Better path:
- fix the source or the materializer that owns the publication
Treating jhf-docs as the owner of product runtime behavior
Problem:
- documentation work becomes a false runtime fix.
Better path:
- use
jhf-docsfor reviewed public content, not product runtime ownership