Automate tenant email replies with human review

A landlord used Claude to read tenant email, revise a lease, and prepare a reply, while still checking everything before it went out.

Done with Claude· Advanced· Jan 2025

What they did
A remote landlord wired Claude Code into WSL2 and connected it to a small Python Gmail command tool that could search threads, read messages, and send replies from the shell. Claude looked up the tenant’s email chain, reviewed the conversation, opened a Markdown lease, changed two requested clauses, turned the file into a PDF with Pandoc, and sent it back. The landlord later said they still verified the work.
How it went
The whole chain took about two minutes and replaced a task that normally took them days. The main issue was trust: commenters objected to having AI edit and send legal contracts, and the landlord later emphasized that a person still needed to review the output.
Tips
This setup depended on a custom Gmail CLI of roughly 200 lines and on storing contracts in Markdown so Pandoc could make the PDF. If you try something similar, keep a human approval step before any email is sent or any contract is finalized.

Try it yourself with Claude

I manage rentals at [property]. Read the tenant’s latest message: [paste message]. Check the email thread, draft a calm reply that matches my lease terms [paste key terms] and local rules in [location], and flag anything that may need formal notice or legal advice. If a lease change is needed, edit this Markdown lease text: [paste lease text] and prepare the revised version for review, but pause and ask for my approval before sending anything or producing a final file. Also write a short note for my records summarizing what changed and why.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on reddit.com ↗

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.