MCP contact enrichment reference

Contact enrichment processes contact rows and can consume credits. Confirm the input, requested fields, quote, and output behavior before starting.

For AI agents

Verify the source rows and available contact fields first.
Show the quote and wait for approval before enrich_contacts.
Preserve the returned run identifier and use get_workflow_run for asynchronous completion.
Do not paste credentials or untrusted contact data into unrelated systems.

Enrich contacts

Enrich the contacts in dataset ds-7f3c2a with work email and phone. Show the quote first and wait for approval.

{
  "account_id": "5093",
  "dataset_id": "ds-7f3c2a",
  "fields": ["work_email", "phone"],
  "confirm": true
}

Call enrich_contacts only after approval. If it returns a workflow run, call get_workflow_run with the same account and run ID. A timeout does not justify submitting the enrichment again.

Input requirements

Use a dataset with contact anchors such as a LinkedIn URL or a matched person record. For a broad uploaded dataset, match or enrich the company context first so contact enrichment has enough information to resolve each row. Before selecting fields, call get_dataset_fields and use the exact field names returned for the current dataset. The returned schema is authoritative; do not copy field names from another dataset without checking them first. enrich_contacts is asynchronous. Preserve the returned workflow_run_id, poll with get_workflow_run, and use the completed output dataset ID. If a wait expires or the connection drops, recover that run before trying again to avoid duplicate paid submissions. See MCP limits and credits and MCP security best practices.

Get Started

Overview

Tutorials

Guides

Reference

Support

MCP contact enrichment reference

For AI agents

Enrich contacts

Input requirements

​For AI agents

​Enrich contacts

​Input requirements

For AI agents

Enrich contacts

Input requirements