Available on Pro

Give AI agents clean access to the property data you already collected.

Connect through MCP or a versioned REST API. Agents can inspect existing Cloud jobs, read normalized results, and create machine-readable exports—without permission to run billable jobs or change your account.

Connect with MCPRead the integration guide

Included with Pro, Ultra, and eligible legacy Team and Enterprise plans.

Agent tool call
read-only
pdl_list_properties({
  runId: "pdl_run_importer_…",
  limit: 25
})

→ 25 normalized records
→ nextCursor: "eyJ2Ijox…"
→ no Cloud credits used

A bounded path from Cloud result to agent workflow.

The integration is designed for deliberate access, compact context, and predictable machine output.

01

Grant only the access it needs

Create a separate credential for each agent and choose scopes for jobs, results, and exports.

02

Let the agent inspect your data

Use MCP tools or the REST API to list existing jobs, page through normalized property records, and read saved insights.

03

Move larger results into a file

Create a versioned JSON or NDJSON export instead of pushing a full property dataset into the model context window.

Designed for controlled access

Useful to the agent. Legible to the person who granted access.

Every credential is scoped, revocable, and separate from your browser-extension login. Large datasets move through expiring export files instead of oversized tool responses.

Access boundaries

  • Reads only the Cloud jobs and results owned by your account
  • Cannot create, edit, start, archive, or delete Cloud jobs
  • Does not spend Cloud collection credits for reads or agent exports
  • Uses normalized fields with explicit provenance, timestamps, units, and null values

Bring your PropertyDataLabs results into the agent workflow you choose.

Create a scoped token, connect your MCP client, and keep complete control over the Cloud jobs and result operations your agent can use.

Connect with MCP