← Back to TeamOpenClaw.com Get the AI Team Guide

Agent profile

Research Agent

The Research Agent keeps you plugged into your market without spending hours in MLS dashboards. It compiles the weekly numbers solo agents need—pricing trends, days on market, inventory shifts, migration data—and packages everything as structured JSON for the downstream agents that write copy, scripts, and offers.

Primary outputWeekly Market Report JSON
TriggerAutomated cron · Monday 6:00 AM
Human oversight5-minute skim + approval

What the Research Agent Delivers

Every Monday morning the agent runs its playbook: pull the latest MLS snapshot, cross-check it with trusted open data (Redfin, Census, migration feeds), normalize the numbers, and log the narrative you need to steer the week. It pushes one JSON payload and one human-readable summary so the rest of the AI team can work without pestering you for clarification.

Inputs

  • MLS export (CSV/IDX) for the target county or city
  • Redfin Data Center API (price per square foot + migration)
  • Inventory + DOM from your brokerage dashboard
  • Optional: macro indicators (mortgage rates, unemployment)

Outputs

  • Structured JSON with 20+ metrics, YoY + MoM deltas
  • Plain-English synopsis (2–3 paragraphs)
  • Flag list of “watch items” (price drops, hot ZIPs, new builders)
  • Audit trail saved to /mission-control/public/data/reports/

Downstream agents

  • Content Strategy Agent ingests the JSON to plan posts
  • Script Agent pulls the hooks + stats for long/short form
  • Lead Gen Agent uses the highlights to craft offers

Implementation Checklist

  1. Connect data sources: Drop your MLS CSV in the shared Dropbox/Spaces bucket or wire the RETS credentials to OpenClaw’s secure secret store. Add Redfin + Census endpoints in mission-control/src/data/sources.json.
  2. Define the schema: Use the provided Research Agent prompt (Chapter 5) and specify the exact JSON keys your other agents expect (medianPrice, dom, newListings, etc.).
  3. Schedule the job: Add a cron entry via Mission Control or openclaw cron add (Monday 06:00). Pipe the log to logs/research.log so you can review fails.
  4. Review + approve: Spend five minutes scanning the summary each Monday. If anything looks off, tag the agent with “rerun” and it will regenerate with fresh pulls.

When to Deploy

  • Weekly cadence before your content/calendar meeting
  • Day-of listing presentations (rerun with ZIP filters)
  • Market volatility (rate hikes, sudden inventory spikes)

Guardrails

  • Never publish the raw CSV—only derived metrics
  • Log the source + timestamp in reports/index.json
  • Alert the human if a data source fails two runs in a row

Ready to install the Research Agent?

The full prompt, JSON schema, and cron template live in Chapter 5 of “Your AI Real Estate Team.” Grab the guide, follow the wiring diagram, and the Research Agent will be shipping reports before your first coffee.

Get the AI Team Guide