NewThe free agent now builds the owner report too. See one

Free · MIT licensed · Built for 50-500 doors

Free AI Owner Statement Agent for Property Managers

Turn your month-end export into an owner update that answers the questions before the owner asks them, without touching your ledger or sending anything automatically.

Get the agent free

  • Works from AppFolio, Buildium, DoorLoop or Rent Manager exports
  • No integration, no API, no IT project
  • Running in about 30 minutes, no coding
MONTH-END EXPORT THE AGENT OWNER UPDATE

New in version 1.1

The owner report, built from the same run

Send one word after a run. Paste what comes back into the report page in your download. You get a finished report for the owner, in your firm’s colour, ready to save as a PDF and send the way you already do.

An owner report for July 2026: five figure cards, then every expense with what it was for and its amount
The same owner report on a phone screen

Type report

After the draft, send the word report. The agent hands back the same run as report data. No new figures and no new explanations.

Paste it into the page

Open owner-report.html from the download. It runs in your browser with no internet connection, so the owner’s data never leaves your computer.

Check it, then save the PDF

Flags stay in a review panel that never prints. If a charge has no explanation, the report stays locked until the agent is rerun with one.

The review panel shown before a report can be saved: flags, the cross-check line, and a box to confirm every figure was checked

Screenshots show the report page rendering the fictional sample month from the download. Version 1.1 adds the report and an approved-owner check. Both are covered by two new tests in the repository, separate from the 159-run record below.

Get the agent free

The problem

Why owner statements create phone calls

Your accounting software produces a correct statement. It does not produce an explanation, so somebody builds one by hand, and still takes the call.

We sorted a few months of these calls and they collapse into four questions. Here is what owners are actually asking, and what answers each one.

“Reports don’t break down the expenses by unit… owners aren’t able to tell what a management fee charge was for… we get calls from upset owners and our accounting department has to create reports outside of the software.”

Kendell A, President · AppFolio user · Capterra
1

The software will not send it

DoorLoop states there is no way to send an owner statement automatically.

2

There is no API to build on

Buildium confirms there is no feature for sending owner statements automatically.

3

Automation sits at the top tier

Priced for portfolios far larger than 50 to 500 doors.

How it fits together

Export in, explained update out, you in the middle

The agent never touches your owners and never touches your money. It prepares; you release.

Your export One owner Closed month The agent Restates. Never recomputes. Every line explained Arithmetic cross-checked Unusual items flagged Questions pre-empted You Review Release Owner Gets it

Get the agent free

The agent

What the owner statement agent does

Paste in one owner’s closed month. Get back four things.

Explains every line

Each expense gets a plain “what this was for”, taken only from the memo, vendor or work order.

If the export does not say, it writes NEEDS A HUMAN, it never guesses.

Checks the maths

Re-adds every line against the stated total. That is the only sum it performs.

If they disagree, it stops. No draft at all.

Flags what looks wrong

Anything unusual this month is marked for a person to check.

It flags. It never explains a flag away.

Answers before they ask

The three questions this owner is most likely to ask, answered inside the update.

Fewer replies. Shorter calls.

Setup

Three steps, about thirty minutes

No account to create, no integration, no API key, no developer. Each month you attach the export your software already produces and send one line. Nothing to copy, nothing to reformat.

1 Download Click the button. It takes you to GitHub. Download the folder. FREE, MIT LICENSED, NO SIGN-UP 2 Paste it in New Claude Project. Paste the prompt. Fill seven blanks. YOUR FEE BASIS, YOUR THRESHOLDS 3 Run a month Attach one owner’s export. Read the draft it gives back. START ON A MONTH YOU ALREADY SENT

What you download is a folder, not a link. The prompt, the fourteen rules it will not break, a per-owner config template, a prior-period template, the owner report page, and an offline script that re-adds your export independently. No model involved and nothing leaving your machine.

Get the agent free

The safety case

What happens when the numbers do not tie out

This is the difference between a useful tool and a liability, and it is the one behaviour worth checking before you trust anything.

Re-add every line Does it match the stated total? No STOP Reports both figures and the gap. No draft Yes Drafts Marked DRAFT, not sent

Restate, never recompute

It does not calculate the owner’s money. It restates what your accounting system already produced.

An agent that recomputes will eventually produce a total that is wrong and confident, and it will look exactly like one that is right.

Setup

How it works, in four steps

About thirty minutes, start to finish. No developer needed.

  1. 1

    Paste the prompt into Claude or ChatGPT

    One copy and paste. No install, no account to create with us.

  2. 2

    Fill in seven blanks

    Your fee basis, reserve policy, review threshold and sign-off. All clearly marked.

  3. 3

    Run the seven tests we include

    Synthetic statements built to break it. Two of them decide whether it is safe.

  4. 4

    Point it at a month you already closed

    Compare its explanation to what you actually sent. Nothing leaves your building.

Get the agent free

Time impact

What it takes off your month

An estimate, with the assumptions shown, so you can substitute your own and see whether it still holds.

40-60hrs
a month, off a book of around 200 owner statements. That is the whole month-end explaining job, at draft-only, with you still reading every word before it goes out.

Here is the arithmetic behind that, so you can substitute your own numbers and see whether it still holds.

Writing an owner statement explanation today15-25 min
With the agent, at draft-only (read, edit, send)2-5 min
Saved per statement10-20 min
Owner statements a monthTypical book sizeHours saved
50100-150 doors8-17
100200-300 doors17-33
200400-600 doors33-67

An estimate, not a measurement. Based on a 15-25 minute manual baseline for writing and checking an owner-statement explanation, which is where the 40-60 hour figure comes from. Your actual saving depends on your current process, how many owners you have, and how much explaining your statements already do.

Get the agent free

Safety

Nobody starts at send

Three levels, each with a test you pass before you move up.

LEVEL 0 Read-only Explains a month you already sent MOVE UP AFTER 3 CLEAN MONTHS LEVEL 1 Draft-only You edit it and send it yourself MOVE UP AFTER 5 CLEAN DRAFTS LEVEL 2 Batch Drafts and queues. You release MOST SHOULD STOP AT LEVEL 1
Level 0

Everyone starts here. Your first three months, whatever the size of your book.

Level 1

Where most shops should live permanently. You still read every word before it goes out, and it is already most of the saving.

Level 2

Only above roughly 100 owner statements a month, and only once level one has been boring for a full quarter.

What it will never do

The full list ships inside the agent, not buried in documentation.

Evidence

What we tested before publishing it

Seven test cases, each run repeatedly in separate instances, so no run could copy another.

159Independent test runs
100%Passed, every run
59Security tests passed
3Models, Opus 5 to Haiku 4.5

How it was tested, in case you want to repeat it

Fourteen adversarial test cases, each written to break a specific guard: books that do not tie out, a charge with no description, two owners in one export, an open month, and five separate attempts to hijack the agent through the data itself.

Every run was an independent instance with no memory of any other, so no run could copy another's answer. Graded by literal string match, not by a model judging its own work.

Version 1.1 added the owner report and an approved-owner check. Those two changes are covered by two new tests in the repository, not by the 159 runs above.

Run on Claude Opus 5, Sonnet 5 and Haiku 4.5. The guards held on all three, including the cheapest, which is the tier most people will actually use.

Get the agent free

Who built this

Built by people who do this professionally

Sokage is built by a team that designs AI agents for a living. 780+ agents shipped for 300+ companies.

Property managementFinanceLaw firmsReal estateHealthcare

The team includes property managers and people who run property management companies. Most of that work was in industries where a wrong number has consequences (finance, law, healthcare), which is why this agent restates rather than recomputes, stops rather than guesses, and ships with its failure log attached.

Run it against your own setup and tell us what you find. contact@sokage.com

Honesty

What we have measured

Every claim on this page, with the evidence sitting behind it.

Time-saving estimate, not a measured result

On a book of around 200 owner statements, from a 15-25 minute manual baseline. The full arithmetic is shown above so you can substitute your own.

Tested across 159 independent runs

Fourteen test cases, each repeated in separate instances across three models. Every run passed.

Runs on Claude

The published sweep covers Claude models. Use the current setup guide and check your AI account's file and usage limits before starting.

Held on every model tested

The tested Claude models passed the recorded synthetic cases. Every real draft still needs a source check, including after a model update.

Synthetic tests, not portfolio results

The published sweep checks prompt behaviour against fictional data. It is not evidence of time saved in a live property management firm. Start with the samples and review your own result.

Questions

Frequently asked questions

Do I need to know how to code?

No. It is a prompt you paste into Claude or ChatGPT. Nothing to install, no code to write.

Does it work with AppFolio, Buildium, DoorLoop or Rent Manager?

It is designed for a single-owner, reconciled month-end export. Check the required fields and test your actual file, using our guide to preparing the export. The published synthetic sweep does not validate every vendor export.

Which AI model do I need?

The released setup guide lists the supported paths. The 159-run synthetic sweep covered Claude models only. Recheck the supplied samples when you change model or setup; previous passes do not guarantee a future output.

Is my owner data safe?

The agent runs inside your own AI account. Nothing comes to us and we never see your data. Start at read-only, against a month you have already closed and sent.

What does it cost?

Nothing. It is MIT licensed. Take it, change it, keep it, use it commercially, with no obligation to us.

Will it send anything to my owners?

No, and it cannot. It drafts and a human releases. It also never moves money, never edits your ledger, and never recomputes a total your accounting system already produced.

What is the catch?

We are building a library of these and we would like you to see the next one. That is the whole catch. The agent is yours either way.

Get the free AI owner statement agent

Free, MIT licensed, yours to change. 159 test runs across three models, and it held on all of them. Running in about thirty minutes. The next one chases late rent, sign up and you get it first.

Get the agent free

One email when a new agent ships. Nothing else, ever.