You have the list. Now someone has to clean it up, work through the properties and decide what deserves attention. LotLine Agent lets you hand that work to the AI you already use.
The instruction goes into your AI client. The work shows up in LotLine, ready for you to review.
> Work my 700-row CSV. Analyze the strongest 150, look up 40 owners, and give me the 15 to review first.
lotline - import_recordslotline - run_batch_autopilotlotline - get_batch_statusBatch complete. Found 36 of 40 owners; four misses weren't charged.
lotline - list_review_queueThe 15 to work first, by score:
...and 12 more. Seller drafts are waiting in LotLine. Nothing was sent.
Illustrative conversation and original LotLine demo capture. Fictional properties; elapsed batch time condensed. Example information is not a property claim, appraisal, bid, or guaranteed result.
A useful request sounds like something you would hand to a teammate: work this list, stay within this budget, and bring me the properties worth a closer look.
Read the guideA list to work. A property to understand. A draft waiting for your review. These examples follow the client treatments in our launch films.
Which of my new leads should I look at first?
These three scored highest in the Harlow County import:
What puts Ashlar St on top?
Its distress score is 78. The owner lives elsewhere, and the county lists an open code violation.
Ask a follow-up about the actual property. The conversation stays connected to the records in your workspace.
CodexLotLine connected> Work the 150 properties I selected. Run Batch Deal Autopilot and tell me which ones need attention.
Called lotline.run_batch_autopilotCalled lotline.get_batch_status148 finished. Two need another look before I can call this done.
I'll inspect those two results before retrying.
The agent can start a batch and check its progress. Long-running work stays visible, including the records that failed.
Hermes AgentLook at the scores first. Use owner lookup on the strongest 40, and stop if I reach my allowance.
Found 36 of the 40 owners. The four lookups that found no one were not charged. The results are saved with the leads.
mcp_lotline_list_leadsmcp_lotline_lookup_ownerChoose where to spend your lookup allowance. Connecting an agent does not change the amount included in your plan.
ChatGPTWalk me through the ARV and rehab estimate for Brackenmere St. What should I verify before making an offer?
Start with the roof and interior condition. The current estimate is based on limited evidence. I would review those assumptions alongside the comps before applying the numbers.
Use the conversation to understand the work already in LotLine. Your comps, inspection and offer decision still need your judgment.
OpenClawWhat seller drafts are waiting for me in LotLine?
22 drafts are ready to review. Want to see the one for Marisol?
Yes, show me hers.
The draft is waiting in Seller CRM with the property and contact details. Nothing has been sent.
With a configured OpenClaw connection, a message in Telegram can bring you back up to speed before you return to the workspace.
5 examples
Illustrative examples · Client appearance and availability vary.
A CSV with 700 properties can be a good starting point. It can also be several evenings of cleanup, scoring, lookups and checking which jobs finished. The work is necessary, but much of it is coordination.
LotLine Agent gives a compatible AI client tools inside your LotLine workspace. You describe the outcome. The client can read leads, start supported jobs, check results and prepare the next step against the same property records you use in the app.
For example: import the list, pick the 150 most promising properties for Batch Deal Autopilot, look up the strongest 40 owners, then bring back 15 to review. That is the sequence shown in the example above. It is a sample workload, not an included monthly allowance or a promise that the batch finishes instantly.
MCP stands for Model Context Protocol. In practice, it lets your AI client ask LotLine to do a supported task and read the result. You connect the client to your workspace, then give it instructions in the conversation.
The connection documentation covers Claude, Claude Code, ChatGPT, Codex, Hermes Agent and OpenClaw. Other compatible clients, including Cursor and Windsurf, may also work. The available connection method depends on the client, its authentication support and your account plan.
The AI chooses from the tools LotLine makes available. LotLine checks access and usage limits on the server. The connection does not give an AI unrestricted control of your browser, other customers' data or every action in the app.
A useful instruction names the list, the result you want and the limits. "Work these 150 properties, look up at most 40 owners, and show me why you picked the top 15" is easier to evaluate than "find me good deals."
Say what matters in your market. You might prioritize recent open code cases, owners with out-of-state mailing addresses, or properties that meet your buy box. Ask the agent to explain its shortlist and separate missing evidence from completed work.
Large batches run as jobs. The agent can check their status and identify records that need attention. A request to prepare tomorrow's queue is a task in that conversation; it does not create a daily schedule by itself.
Seller outreach is a good example of the handoff. The agent can prepare a draft against a lead, and you can review it in Seller CRM with the seller details and the applicable outreach checks.
A prepared draft has not been sent. Agent does not automatically send seller texts, emails or place calls. Review the recipient, the wording and the permission to contact them before using a communication channel.
The same care applies to deal analysis. An ARV or rehab estimate gives you something to investigate. It does not replace current comps, an inspection or the decision about what you are willing to offer.
LotLine Agent is a $49/month add-on for eligible paid Growth Cloud and Team Cloud workspaces. It adds the connection that lets your AI coordinate supported LotLine work. An external AI subscription, when required by that client, is separate.
Owner lookups and other metered actions still use the relevant workspace allowances. Agent does not add a second pool of SMS, voice or lookup credits. You can tell it to work selectively, and server-side limits still apply.
The practical question is how much time you spend moving between steps. If you regularly work batches, check jobs and assemble a shortlist for an acquisition operator, that coordination can be worth delegating. If you only handle a few leads a month, the normal workspace may be all you need.
For the wider cost comparison, see our guide to affordable wholesaling software.
CSV imports, property analysis, seller drafts and the rest of the normal workspace remain available according to your plan. You do not need Agent to click through those workflows yourself.
An operator or VA can use the same property records after the agent has worked a list. The useful handoff is a clear shortlist, the reasons behind it, and the exceptions that need a person. You should not have to reconstruct the job from a wall of chat.
That is also why the difference between a CRM and a deal workspace matters: the property record is where the work stays, whoever starts it.
LotLine Leads is the separate product for building a batch from supported public records. You choose the market and criteria; Leads checks the sources, resolves properties and owners, and preserves the evidence behind the result.
You can use Leads through the web workspace where it is enabled without buying Agent. With both products, a connected AI can help coordinate the supported build and then work the resulting leads. Lead credits and workspace allowances remain separate.
The distinction is simple: Leads builds the pipeline. Agent works the pipeline.
It is an AI client that can use connected tools to carry out supported steps in a real estate workflow. With LotLine Agent, those steps include reading leads, starting analysis, checking jobs and preparing drafts in your workspace.
The connection documentation covers Claude, Claude Code, ChatGPT, Codex, Hermes Agent and OpenClaw, as well as compatible HTTP MCP clients. Client plan, transport and authentication requirements vary.
The add-on is $49/month for eligible paid Growth Cloud and Team Cloud workspaces. Any required external AI subscription is separate, and normal workspace usage limits still apply.
No. It can prepare seller outreach for review. It does not automatically send seller texts, emails or place calls.
It can coordinate supported imports and batch jobs. Your request must fit workspace permissions, batch rules and available allowances. Large jobs take time and may return records that need attention.
No. The normal workspace remains available according to your plan. Agent adds another way to coordinate supported work.
No. Leads can be used in the web workspace where enabled. Agent is optional for coordinating supported Leads workflows through a connected AI client.
It can take on some of the software coordination. A person still needs to assess the property, verify assumptions, handle relationships and make consequential deal decisions.
Product details reviewed October 6, 2026. Check your workspace for current access and allowances.
Choose a small list, give your AI a clear brief, and review the result in LotLine. Check Agent access and connection requirements in your workspace.