In your software
Onboarding · customer 4821
Still missing
- Legal company name
- VAT number
- Registration document
Create an external request
APIs for getting input, files, and approvals from people.
Software and AI agents create a hosted request. Fangst returns a link. The person responds, and Fangst sends the structured result back through the API or a webhook.
Your software creates a request. A person fills it in. Fangst sends the response back.
Fangst hosts the request page, validates the fields and files, stores the accepted response and sends it to your software.
Hosted request
Requested by Ledgerline
Complete your company profile
We need these details before setup can continue.
Back in your software
submission.completed
- customer_id
- cus_4821
- fields
- 2 received
- attachments
- 1 received
signature verified
Ask for one answer, a complete form, files, or an approval.
A Fangst request belongs to a real customer, claim, order or ticket. It can ask for one missing field, supporting documents and photos, a complete onboarding, claim or project brief, or a decision about a specific proposal. Your system keeps its private context. The response returns with the correlation metadata your software supplied.
- One answerA missing VAT number, confirmation or date
- Documents and photosReceipts, damage photos or supporting files
- Complete intakeAn onboarding, claim or project brief
- ApprovalApprove, reject, acknowledge or choose
- ReviewConfirm or correct generated and extracted data
- Human taskComplete an external action and report the result
Fangst handles the request from link to delivery.
Your software defines the input, files, or decision it needs. Fangst hosts the public request, validates the response, stores it durably and delivers it back.
Request infrastructure Fangst handles
- Public access A public upload endpoint, private or expiring links, optional passcodes, rate limits and abuse controls.
- Input and files Server validation, file type and size limits, durable storage and authenticated downloads.
- Data lifecycle Encrypted response storage, API deletion, short automatic retention and coordinated expiry for data, files and backups.
- Delivery failures Signed webhooks, retries with backoff, delivery history and replay after a final failure.
Your software receives the response. Fangst handles the access, storage and delivery around it.
An AI agent can create the request itself.
Give an AI agent a Fangst API token and permission to make HTTP
requests. When it needs information, files, or a decision from a
person, it can create a request and receive the hosted URL. The agent or your
software sends that link through a channel it can use; Fangst does
not send the message. The agent can poll the request submissions,
or your application can continue from the
submission.completed webhook.
- Missing input found The run needs a document or answer from a person.
- Create through the API Fangst returns a hosted URL.
- Send the link The agent uses email, chat or another available channel.
-
Read the response
Poll submissions or handle
submission.completed.
Fangst validates, stores and delivers every accepted response.
A request can be as simple or detailed as the job requires. Its response and delivery still have explicit rules.
- Before response
- Schema snapshot, private URL, optional passcode, expiry and one-shot limits.
- At response
- Server-side validation, typed fields, configured file limits and a receipt.
- After response
- Encrypted storage before HMAC-signed delivery, with retries and short scheduled deletion after delivery.
Fangst stores each accepted response before delivery. Your application remains the system of record and makes every decision afterward.
Give your software a way to ask people.
Your software sends the private link through its own channel. Fangst hosts the request, stores the response and sends it through a signed, retryable webhook.