Complete the Finch Connect flow for each employer to obtain a provider-specific access token scoped to the required data permissions.
Call GET /employer/directory to retrieve the list of individuals and their Finch-assigned individual_ids.
POST to /employer/individual with a JSON body containing an array of individual_ids to retrieve personal data for those individuals in a single batch call.
POST to /employer/employment with a similar array of individual_ids to retrieve employment details such as start date, employment type, income, and location.
Handle per-item errors within the batch response array alongside successful records — Finch may return a mix of successes and per-ID errors in a single response.
Known gotchas
There is no batch size cap on the /individual or /employment endpoints — send all required IDs in one request rather than artificially splitting into smaller chunks.
Different providers expose different subsets of fields; a field present for one provider may be null or absent for another — always code defensively against missing attributes rather than assuming field presence.
Finch access tokens are per employer per provider; an integration connecting multiple employers requires a separate Connect flow and token for each employer.
Give your agent this knowledge — and 200+ more routes
One MCP install gives any agent live access to the full route map, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp