The moment it happens

Pull details out of PDFs and photos

Send a document link and get back the fields you asked for, with personal ID numbers masked.

Runs: The moment it happens
Apps: Anthropic API
AI: Not needed
Access: API key

Send this to your assistant

Create a ClickRuntime automation: when I post a PDF link to this webhook, extract the applicant's name, income and employer, score their loan eligibility from 1 to 5, mask any ID numbers, and email me the result.

Paste it into your assistant. It tells you what it will send or change before it runs live.

Add to ChatGPT Add to Claude

First time? Adding ClickRuntime to your assistant takes about two minutes.

Why use it

Why teams use this automation

Teams receive documents that need reading before anything else can happen: supplier invoices, signed forms, receipts, ID photos, damaged-goods pictures sent by a customer. Typing the contents into a system is slow and error-prone, and traditional OCR struggles with varied layouts, handwriting and photos taken at an angle. A claims handler or an accounts clerk spends part of every day squinting at attachments and copying numbers into a spreadsheet, and mistakes are only found when a payment goes to the wrong amount.

When a file link arrives, the automation downloads the file and sends it to the Anthropic API as a document or image content block along with your instructions. Claude reads it and returns the fields you asked for, such as vendor, total and due date, or a verdict like undamaged or damaged. The result is emailed to you.

Finance teams, claims and support staff, and anyone who handles incoming paperwork or photos can use the same pattern with different instructions.

What you get out of it

Data entry from documents falls sharply
Reading an invoice or a form and extracting the fields takes seconds, so staff check the result instead of typing it, and a day's backlog clears in minutes. A person still reviews the result, but the typing is gone.
Varied layouts are handled
Claude reads documents by understanding them, not by matching fixed positions. Invoices from new suppliers or handwritten notes do not need a new template before they can be processed. New suppliers do not slow you down.
Photos can be assessed on the spot
A customer's picture of a damaged parcel can be checked against your criteria straight away, so the claim is categorized and answered in minutes, not the next working day. Customers get a quicker answer.
Structured output feeds other tools
Because the answer comes back as clear fields, it can go straight into a spreadsheet, accounting tool or database with no retyping, and mistakes from manual entry disappear. It also shortens the path from paperwork to payment.
A consistent standard for every document
The same instructions are applied to every file, so a verdict does not depend on who was on shift. That helps with audits and with explaining decisions to customers. That makes outcomes easier to defend.

How it works

What happens, step by step

When you send a document link, such as a loan application PDF, to your personal ClickRuntime link, the automation starts straight away.
Claude reads the document and pulls out the applicant's name, income and employer.
Claude also scores loan eligibility from 1 to 5 and explains why.
Long ID numbers are masked so only the last four digits show, and the results are emailed to you.

Before it runs live

Connect Anthropic API
Paste your automation's link into the other app's webhook settings

What you get

RunsThe moment it happens

Pull details out of PDFs and photosjust nowSucceeded

Email · Loan applicationsLoan application: Dana Reyes scored 4/5 {"income": 86000, "employer": "Northwind Traders"}

Build your own

Need it to work differently?

This template is a starting point. Copy the prompt and change it, or describe a different automation from scratch, with your own conditions, more apps or an AI step. Your assistant writes it, tests it and runs it on ClickRuntime.

Your own rules
Conditions, thresholds, filters and exceptions, written the way you'd explain them to a colleague.
Any number of apps
One automation can read from one app and update several others. Over 248 apps are ready, and anything with an API works too.
AI where it helps
Summaries, classification, extraction and drafts use ClickRuntime's built-in AI, with no extra key.
On a schedule or on each event
Runs as often as every 5 minutes, or the moment an app sends an event, and keeps track of what it already handled.

Changes you can ask for

Act only on the items that match your own rules, and skip the rest
Add an AI step, for example to summarize, sort or draft a reply
Add another app, for example also send it to Slack
Combine it with other Anthropic API steps into one automation

Under the hood

The code that runs this automation

This is the Python script that runs each time the automation starts. Your AI assistant writes it for you when you send the prompt, and you allow each step in your assistant. To change how the automation works, describe the change to your assistant and it updates the script for you. For every function it can use, see the script reference.

analyse_pdfs_and_images_with_claude.pyexample

import json, re
from clickruntime import on_event, credentials, notify, http

def claude(model, prompt, **extra):
    resp = http.post("https://api.anthropic.com/v1/messages",
                     headers={"x-api-key": credentials.secret("ANTHROPIC_API_KEY"), "anthropic-version": "2023-06-01"},
                     json={"model": model, "max_tokens": 1024, "messages": [{"role": "user", "content": prompt}], **extra},
                     timeout=120)
    resp.raise_for_status()
    return resp.json()

@on_event
def handle(event):
    url = (event.json or {})["url"]
    msg = claude("claude-sonnet-5-5", [
        {"type": "document", "source": {"type": "url", "url": url}},
        {"type": "text", "text": "Extract the applicant's name, income and employer, and score loan eligibility "
                                 "1-5. Reply with JSON only: {\"name\", \"income\", \"employer\", \"score\", \"reason\"}."},
    ])
    result = json.loads(msg["content"][0]["text"])
    body = re.sub(r"\b\d{6,}\b", lambda m: "*" * (len(m[0]) - 4) + m[0][-4:], json.dumps(result, indent=2))
    notify.email_me(f"Loan application: {result['name']} scored {result['score']}/5", body)
    return {"score": result["score"]}

FAQ

Common questions

Which file types can it read?

PDFs and common image formats such as JPEG, PNG, GIF and WebP, within the size limits of the Anthropic API. Very large PDFs or very high-resolution images may need to be split or reduced first.

Does this need my own Anthropic API key?

Yes. The call goes to the Anthropic API directly, so your key is saved as a secret. Usage is billed by Anthropic to your account according to the model you choose. Costs depend on the model and the size of the files.

How do I know the extracted figures are right?

Ask for the source text or a confidence note next to each field, and have totals cross-checked, for instance that line items sum to the invoice total. Anything that fails the check can be flagged for a person.

What permissions does it need?

API key. Nothing else in your accounts is used.

Does it need AI?

No. It runs without AI, so it doesn't use your plan's AI allowance.

How many runs does it use?

One run for each event the app sends.

Can I change how it works?

Yes. Describe the change to your AI assistant, for example a different schedule, channel or filter. It updates the automation, tests it and tells you what changed.

The moment it happens