I built a lease agreement analyzer in n8n — extracts rent, deposits, and red flags, then generates tenant negotiation advice before you sign

Built with n8n
1

Most people sign a lease after skimming it for 10 minutes. The auto-renewal clause on page 6, the early termination penalty buried in section 12, the vague maintenance language that lets a landlord argue anything — you don’t catch those in a skim. Built a workflow that reads the whole thing and tells you what to watch out for.

image
image1236×435 86 KB

What it does

Lease PDF uploaded to Drive → runs two AI passes → calculates move-in cost → logs to comparison spreadsheet → routes to high-risk alert or standard notification on Slack

Takes about 25-30 seconds per lease.

Pass 1 - Structured extraction:

Pulls every financial term, date, policy, and clause out of the lease into structured fields.

Pass 2 - Tenant advice:

Reviews the same lease from a tenant advocate perspective. Writes a 2-3 sentence summary, lists the top 3 things to negotiate before signing, rates the lease as Tenant-Friendly / Neutral / Landlord-Friendly, and flags anything that could cause problems at move-out.

What gets extracted

Financial terms:

  • Monthly rent, security deposit, pet deposit

  • Total move-in cost (rent + security + pet deposit combined)

  • Late payment grace period and fee

  • Returned check fee

  • Early termination penalty

  • Parking fee

Dates and duration:

  • Lease start and end dates, term length

  • Auto-renewal (yes/no) and renewal terms

  • Required notice to vacate (days)

  • Move-out notice required (days)

Policies:

  • Pet policy — allowed or not, deposit amount, monthly pet rent, breed/size restrictions

  • Utilities included — which ones, which are tenant-paid

  • Parking included or fee

Parties:

  • Landlord name, company, phone, email

  • All tenant names listed

Red flags:

Every potentially unfair clause gets flagged with severity (High / Medium / Low) and the actual clause text. Examples the workflow catches:

  • Excessive late fees (over 5-10% of rent)

  • Unreasonable early termination penalties (more than 2 months rent)

  • Vague maintenance language (“tenant responsible for all repairs”)

  • Missing standard protections (no habitability clause)

  • Auto-renewal with inadequate notice requirements

Risk levels

  • High: 2+ high-severity flags → fires a separate urgent Slack alert

  • Medium: 1 high flag or 3+ total flags → standard notification

  • Low: minor or no issues → standard notification

The Slack alerts

High risk lease:


🚨 HIGH RISK LEASE DETECTED

Property: 123 Main St, Apt 4B

Rent: $2,400/month | Move-in: $7,200

Term: 2025-06-01 to 2026-05-31

⚠️ Red Flags (3 found):

[High] Early termination = 4 months rent ($9,600)

[High] Landlord can enter with 12 hours notice (not 24)

[Medium] Tenant responsible for all appliance repairs

📋 File: lease-123-main.pdf

Standard lease:


✅ Lease Analyzed

Property: 456 Oak Ave, Unit 2

Rent: $1,800/month

Term: 2025-07-01 to 2026-06-30

Risk Level: Low

File: oak-ave-lease.pdf

What lands in Google Sheets

Each row: Property Address, Landlord, Tenants, Monthly Rent, Security Deposit, Move-In Cost, Lease Start, Lease End, Pet Policy, Utilities Included, Late Fee, Early Termination, Red Flags, Risk Level, Tenant Advice, Analyzed Date

Useful if you’re comparing multiple apartments — all your leases in one sheet side by side.

Setup

You’ll need:

  • Google Drive and Sheets (free)

  • n8n instance (self-hosted — uses PDF Vector community node)

  • PDF Vector account (free tier: 100 credits/month, roughly 12-15 leases)

  • Slack (optional — delete the last two nodes if not needed)

About 15 minutes to configure.

Download

Workflow JSON:

github.com/khanhduyvt0101/workflows

Full workflow collection:

khanhduyvt0101/workflows

Setup Guide

Step 1: Get your PDF Vector API key

Sign up at https://www.pdfvector.com — free plan works fine. Go to API Keys and generate a key.

Step 2: Create your Google Drive folder

Create a folder called “Leases” in Google Drive. Copy the folder ID from the URL (string after /folders/).

Step 3: Set up your Google Sheet

Create a new spreadsheet with these exact headers in Row 1:


Property Address | Landlord | Tenants | Monthly Rent | Security Deposit | Move-In Cost | Lease Start | Lease End | Pet Policy | Utilities Included | Late Fee | Early Termination | Red Flags | Risk Level | Tenant Advice | Analyzed Date

Copy the Sheet ID from the URL.

Step 4: Import the workflow

Download the JSON from GitHub and import into n8n via Import from File.

Step 5: Configure the nodes

Google Drive Trigger:

  • Connect your Google account

  • Paste your Leases folder ID

Download Lease:

  • Same Google credential

PDF Vector - Analyze Lease:

  • Add new credential (Bearer Token type)

  • Paste your API key

PDF Vector - Tenant Advice:

  • Same PDF Vector credential

Log Lease Analysis:

  • Connect Google Sheets

  • Paste your Sheet ID

  • Sheet tab must be named exactly “Sheet1” (or update to match your tab name)

High Risk Alert (Slack):

  • Connect Slack

  • Paste your channel ID for urgent alerts

Analysis Complete (Slack):

  • Same Slack credential

  • Can use the same channel or a different one

Step 6: Test it

Activate and drop a lease PDF into your Drive folder. Check your Sheet and Slack after about 30 seconds.

Accuracy

Tested on residential leases from the US — standard form leases, custom landlord agreements, and multi-page commercial-style residential leases.

  • Financial terms (rent, deposits, fees): ~97% on digital PDFs

  • Dates and duration: ~95%

  • Pet policy and utilities: ~93%

  • Red flag detection: ~85% — catches obvious issues reliably, may miss unusual jurisdiction-specific clauses

  • Tenant advice quality: good for standard leases, less specific on highly customized agreements

Scanned leases: drops to ~85% overall. Works but worth double-checking the financial figures.

Cost

Each lease uses about 6-8 PDF Vector credits (two AI passes). Free tier of 100 credits gets you roughly 12-15 leases per month.

Basic plan is $25/month for 3,000 credits if you’re a property manager running multiple leases regularly.

Customizing it

Adjust risk thresholds:

In the Compile Analysis code node, the risk logic checks highRisks.length >= 2 for High and highRisks.length === 1 || redFlags.length >= 3 for Medium. Change these numbers to be stricter or more lenient.

Add commercial lease support:

Edit the extraction prompt in the PDF Vector - Analyze Lease node to include commercial-specific terms like CAM charges, tenant improvement allowances, and percentage rent clauses.

Skip Slack, use email instead:

Replace both Slack nodes with Gmail nodes. The template variables from $('Compile Analysis').item.json work the same way.

Trigger from email:

If leases come in as email attachments, swap the Drive trigger for a Gmail trigger. Everything downstream stays the same.

Limitations

  • Requires self-hosted n8n (PDF Vector is a community node)

  • Red flag detection is AI judgment — not legal advice, always have an attorney review before signing important leases

  • Jurisdiction-specific tenant protections vary — the AI doesn’t know local law

  • Highly customized or unusual lease formats may extract less cleanly

  • Doesn’t track lease expiration dates or send renewal reminders (would need a separate scheduled workflow)

Links

Questions? Drop a comment if something’s not working or you want to adapt it for commercial leases or property management.

2

clever — legal tech via n8n for end users, not just enterprises. the red flag detection is the core piece: how do you define what counts as a red flag — purely rule-based (specific clauses/keywords) or does the model decide on its own? and how do you handle country-specific legal differences? lease law in DE vs AT vs CH is pretty different.

3

Hey Benjamin - good catch on the jurisdiction issue. This is definitely US-centric right now.

Red flag detection: It’s AI-based, not rule-based. The extraction prompt tells the model to flag “potentially unfair clauses” and categorize severity. So it’s making judgment calls based on what it’s seen in training, not hard-coded keyword matching.

Problem is: “unfair” means different things in DE vs AT vs US. What’s a red flag in California (no 24-hour entry notice) might be totally legal in Germany.

Honest limitation: The model doesn’t know local tenant protection laws. It’s basically applying “generic tenant advocacy” logic, which works okay for standard US residential leases but breaks down fast outside that.

To make this work across jurisdictions, you’d probably need:

  • Country-specific prompts with actual local law references
  • Different red flag thresholds per region
  • Maybe even country-specific extraction schemas

Right now it’s more “sanity check tool” than “legal compliance validator.”

4

the “sanity check not legal compliance” framing is exactly right — useful for flagging things worth reading more carefully, not replacing an attorney. the jurisdiction issue is real though. German lease law specifically has a bunch of quirks that a model trained mostly on US data would likely miss: Schönheitsreparaturen clauses (cosmetic repair obligations) are frequently declared void by courts but still show up in leases, Betriebskostenabrechnung formalities have very specific rules around what landlords can actually charge, and notice periods work differently depending on contract type. would actually be interesting to test how it performs on a standard German residential lease — have you run it on anything outside the US?