We would like StoriiCare to provide secure API access so authorised AI agents, such as our internal care operations agent Denise, can interact with StoriiCare data in real time.

Our goal is not to replace StoriiCare, but to make StoriiCare a live source of truth for care operations, similar to how we have integrated with our eMAR system.

Key use cases:

1.⁠ ⁠Live resident registry feed

- Resident profile details

- Room / bed placement

- Admission status

- Discharge / hospital leave / return status

- Preferred hospital

- Diet / allergies / key care flags

2.⁠ ⁠Operational event feeds

- New admission created

- Resident transferred room

- Resident discharged

- Hospital admission / return

- Care profile updated

- Diet/allergy update

3.⁠ ⁠Safe AI-assisted workflows

- Denise can verify resident identity before routing clinical or operational tasks.

- Denise can detect stale room/status information.

- Denise can reduce manual duplicate entry between StoriiCare, eMAR, kitchen planning, and internal dashboards.

- Denise can provide safer handover by using current StoriiCare data instead of exported CSVs.

4.⁠ ⁠Read/write API where appropriate

Ideally we need:

- Read resident list

- Read individual resident profile

- Read recent updates/events

- Webhooks for resident/profile/status changes

- Optional writeback for non-clinical operational notes or structured updates, subject to permissions

Technical requirements we would need:

•⁠ ⁠REST or GraphQL API

•⁠ ⁠OAuth2 / API key / service-account authentication

•⁠ ⁠Role-based permissions

•⁠ ⁠Webhooks for key events

•⁠ ⁠Pagination and filtering

•⁠ ⁠Stable resident IDs

•⁠ ⁠Updated timestamps for sync

•⁠ ⁠Audit log for API access

•⁠ ⁠Sandbox/test environment

•⁠ ⁠Clear data schema documentation

•⁠ ⁠Rate limits suitable for care-home operational sync

Important safety points:

•⁠ ⁠API access should be permissioned and auditable.

•⁠ ⁠AI agent access can be read-only initially.

•⁠ ⁠Any writeback can be limited to approved fields and logged.

•⁠ ⁠Clinical decisions remain with human nurses/doctors; the AI agent only supports workflow, reconciliation, reminders, and safety checks.

This would allow care homes to build safer, more automated workflows while keeping StoriiCare as the central care record.