Acrual API

What authentication does the Acrual API use?

Acrual uses HTTP bearer token authentication. You pass an Authorization: Bearer <token> header on every request. Through Jentic, the bearer token is stored encrypted in your Jentic One instance and never enters the agent's context - the agent receives a scoped execution handle instead of the raw secret.

Can I create invoices and purchase orders with the Acrual API?

Yes. POST to /v1.0/invoices to issue an invoice and POST to /v1.0/purchase-orders to raise a PO. Both can be linked back to a job and a client record. The spec covers 46 operations in total across jobs, invoices, quotes, purchase orders, time entries, suppliers, and webhooks.

What are the rate limits for the Acrual API?

The OpenAPI spec does not declare explicit rate limits, so check Acrual's developer portal for the current ceiling. Build retries with exponential backoff around 429 responses, and prefer webhook subscriptions on /v1.0/hooks for change notifications instead of polling.

How do I subscribe to invoice events with the Acrual API through Jentic?

Run pip install jentic, search Jentic for 'register an Acrual webhook', load the schema for POST /v1.0/hooks, and execute it with the target URL and event type. Jentic handles the bearer auth via its vault, so the only inputs your agent needs are the callback URL and the event filter.

Does the Acrual API support file attachments on jobs?

Yes. The /v1.0/jobs/{id}/files and /v1.0/jobs/{id}/files/folder endpoints let you upload and organise files inside a job folder, which is how construction crews keep site photos, drawings, and signed variations attached to the job record.

Can I limit what my agent is allowed to do with the Acrual API?

Yes. Because Jentic One is self-hosted, you decide which Acrual operations the agent may call, and your own rules determine the credentials it can use. Many of these endpoints, such as creating an invoice on /v1.0/invoices or registering a job webhook on /v1.0/hooks, take the job, invoice, or hook details in the request body rather than a resource id in the URL, so you scope the agent to just the operations it needs. Reading purchase orders on /v1.0/purchase-orders or time entries on /v1.0/time-entries stays off unless you explicitly add those operations.