AI build guide
Build an internal tool with AI that respects permissions
Internal tools often touch the most sensitive business data. Speed matters, but explicit permissions and reversible operations matter more.
01
Map the workflow
Document inputs, decisions, approvals, exceptions, and owners before automating the happy path.
02
Apply least privilege
Limit data and actions by role on the server; hidden buttons are not access control.
03
Measure the operation
Expose useful status, errors, audit evidence, and recovery paths for the people operating the tool.
04
A production-shaped Xroga prompt
Build an internal approval tool for purchase requests. Add requester, reviewer, finance, and admin roles; record status transitions and comments; require server authorization; preserve an audit trail; export approved records; and test denied access and retry behavior.
05
Evidence to require
Require role tests, audit records, empty/error states, and a recovery path for failed integrations before inviting operators.
06
Before calling it production-ready
Review permissions, data handling, error states, accessibility, mobile behaviour, tests, environment configuration, observability, and rollback. The exact checklist must follow the project rather than a fixed framework template.