Lease Summary Agent · Use case
How can a landlord or a property manager use an AI agent to summarise a lease on its own infrastructure?
A lease summary agent reads one lease and its later deeds, and sets out the term, the rent, the review and break dates, the repair duties and the service charge terms, each with the clause it came from. The property team checks the summary against the lease. The agent gives no legal opinion, sets and proposes no rent, decides nothing about a tenant and sends nothing to one.
Every lease is different.
The dates are found by reading all of it.
- ✗ 01A portfolio holds leases signed over decades, each negotiated on its own terms and changed since by licences, variations and side letters.
- ✗ 02A break date or a rent review missed costs a year's income or more, and the summary a team works from is often older than the last change to the lease.
Where it stops for a person
Always. The property team checks the summary against the lease.Always. It summarises a document, it does not advise: it gives no legal opinion, proposes no rent and no action on a date, and takes no view on a tenant. Whether to review, renew or break is the asset manager's decision, and the lease, not the summary, is what the team relies on.
It sets out the terms with their clauses.
The property team checks them.
Step · 01
A property manager uploads the lease and any later deeds, licences and side letters.
Step · 02
The agent reads them in date order and notes where a later document changes an earlier term.
Step · 03
It sets out the term, the rent, the review and break dates, the repair duties and the service charge terms, each with the clause.
Step · 04
It marks every term it could not read with confidence and every place where two documents disagree.
Step · 05
The property team checks the summary against the lease and decides what goes into the lease record.
Reads from · connectors
The lease and the deeds, licences and side letters that change it, as uploaded files. No tenant account and no operational system.
Acts through · MCP servers
Nothing. It has no tool that changes anything, and it writes to no system.
What you set up
- Choose leases your property team has already summarised by hand, with their later deeds.
- Tell the agent the terms and the order your summary template uses.
- Build the agent in Katonic Studio and test its summaries against the ones your team wrote.
- Publish it to the property team, one building first.
Where it runs
The agent, its knowledge store and the model run on infrastructure you own or control. You choose the model and serve it yourself, so questions, documents and answers are not sent to a public AI service. Leases hold personal and commercial data and stay inside the company.
How you get it today
This page is labelled Use case: the agent is built in Katonic Studio. You describe the job, connect what is named above, add test questions and release the agent once it passes them. When the Catalog holds a tested template for this job, this page will carry the Template label and show how to take it.
Does it set or propose a rent?
No. It sets out what the lease says about rent and review, with the clause. It proposes no figure.
Does it decide anything about a tenant?
No. It reads a lease. It screens, scores and ranks no one, and it reads no tenant account.
Does it write to the tenant?
No. It has no tool that sends anything. A property manager writes to the tenant.