Before you start
A signed-in NodeFox workspace. Use synthetic data and prepare outputs before enabling external writes.
01Define completeness
List the required documents and fields before collecting them. Use an eight-location example with sixteen requirements and one missing site contact. Make the requirement IDs stable so the check is reproducible.
02Assemble evidence
Retrieve supported documents or sample API records and normalize them into one checklist. Preserve the source for each requirement. A model can help interpret documents but should not silently invent missing information.
03Route the gap
Code compares supplied requirement IDs with the required set. Send the missing contact to the onboarding owner. Keep the workflow waiting or produce a follow-up packet rather than marking the account ready.
04Prepare the handoff
After the missing field is supplied and reviewed, prepare a readiness packet. Before creating CRM or service records, check for existing identifiers and define how to avoid duplicates.
Expected result
A complete checklist or an explicit list of missing requirements, with owner and source context.
If the result is wrong
Do not measure completeness by document count alone; a present file may still lack a required field.