Every API endpoint

This page lists every endpoint of the veldhost API, version 1.6. It is generated from the same OpenAPI document the API serves at https://manage.veldhost.eu/api/v1/openapi.json, so it always matches what the API actually accepts.

All paths are relative to https://manage.veldhost.eu/api/v1. Every request needs a personal API token (or an assistant's access token) in the Authorization: Bearer … header; see API tokens and authentication. Endpoints marked Idempotency-Key accept that header, so a retried request is safe; see Errors, rate limits and idempotency.

For request and response fields, load the OpenAPI document into any OpenAPI tool, or read the API reference for the main workflows.

Sites

Hosted sites: state, deploys, logs, health, usage.

Site settings

A site's runtime, build steps, environment variables, redirects and domains.

Releases

CI/CD releases: approve one awaiting approval, discard it, or roll back to an earlier one.

Backups

Restore points and staging copies.

Cron

Scheduled tasks (PHP runtimes).

Domains

Registered domains and the registry.

DNS

Authoritative DNS zones and records.

Email

Domains with veldhost email: their mailboxes, verification and the DNS records they still need.

Data

Databases: your site databases (MariaDB and PostgreSQL) and warehouse databases, their schema, read-only SQL on sites and writes on the warehouse. Needs the opt-in data scope.

Dashboards

Dashboards: grids of charts, stats and tables over your databases, with a time range, auto-refresh and variables. Panel SQL may use the dashboard macros ($__timeFilter, $__timeGroupAlias, $__interval, ${variable}); the server expands them and answers from a short-lived cache. Needs the opt-in data scope.

Pipelines

Pipelines: SQL that builds tables in one of your warehouse databases. Databases are named in the SQL by the name /data/databases shows (shop.orders, or shop.public.orders on PostgreSQL). Two statement forms — CREATE OR REFRESH TABLE [.] [KEY (cols)] AS , built in the warehouse (KEY upserts, no KEY rebuilds), which may read site tables as .; and CREATE OR REFRESH SOURCE VIEW [FROM ] AS , which runs on the site database in its own dialect and stages only the rows it selects. The order tables are built in is derived from the SQL: every name after FROM or JOIN. Compiling is free and answers with a file and a line per problem. Needs the opt-in data scope.

Method	Path	What it does	Idempotency-Key
GET	/pipelines	Your pipelines, every one on your accounts (data scope)
POST	/pipelines	Create a pipeline (data scope; owner, admin or developer on an account with an active plan). Every database it names is checked against YOU now and against its author each time it runs. The answer carries compile, so read it before telling anyone it works	Yes
GET	/pipelines/{id}	One pipeline: its SQL files, its sources, the derived graph and how the last run went
PUT	/pipelines/{id}	Replace the whole document (send the version you read; 409 stale_version when it moved on). Saving makes you the author, and therefore the authority its runs act with. A pipeline whose SQL does not compile still saves — compile.ok says so	Yes
DELETE	/pipelines/{id}	Delete a pipeline and its run history (author, or account owner/admin). The tables it built stay in your warehouse	Yes
POST	/pipelines/{id}/compile	What the SQL means, without running it: every problem with its file and line, the dependency graph, the build order and the plan. Reads nothing and writes nothing, so iterate here rather than on a run
POST	/pipelines/{id}/preview	Try it on a sample without writing anything: a couple of hundred rows per source view into throwaway tables, each table built into a temporary one, a sample of the rows back, everything dropped again	Yes
POST	/pipelines/{id}/run	Build every table now. One run per account at a time: a pipeline asked for while something else is running answers with a run whose status is skipped	Yes
GET	/pipelines/{id}/runs	This pipeline's run history, newest first

Jobs

Jobs: an ordered list of steps we run for you — SQL on a warehouse database, refreshing a dashboard, emailing a report, calling one of your webhooks, waiting, or starting another job — on a cron schedule in your own time zone, after another job, or when you ask. Needs the opt-in data scope.

Method	Path	What it does	Idempotency-Key
GET	/jobs	Your jobs, every one on your accounts (data scope)
POST	/jobs	Create a job (data scope; owner, admin or developer on an account with an active plan). Every database a step names is checked against YOU now and against the job's author each time it runs	Yes
GET	/jobs/{id}	One job: its trigger, steps, budget and last run
PUT	/jobs/{id}	Replace the whole model (send the version you read; 409 stale_version when it moved on). Saving makes you the job's author, and therefore the authority its scheduled runs act with	Yes
DELETE	/jobs/{id}	Delete a job and its run history (author, or account owner/admin)	Yes
POST	/jobs/{id}/run	Run this job now. One run per account at a time: a job that is already running answers with a run whose status is skipped (concurrency skip) or queued (concurrency queue). When the account's monthly spend cap on job minutes is reached the call is refused with 422 error type usage_cap and a billing_url	Yes
GET	/jobs/{id}/runs	This job's run history, newest first
GET	/jobs/runs	Every run on your accounts, newest first — jobs and pipelines in one view
GET	/jobs/runs/{id}	One run and its steps, optionally after waiting up to wait_s for it to settle (200 settled, 202 still going)
POST	/jobs/runs/{id}/cancel	Ask a run to stop. A queued run stops at once; a running one stops between steps, never mid-statement	Yes

Webhooks

Outbound webhooks and chat channels.

Method	Path	What it does	Idempotency-Key
GET	/webhooks	List your outbound webhooks (manage scope)
POST	/webhooks	Add an outbound webhook. The signing secret is returned ONCE in this response	Yes
DELETE	/webhooks/{id}	Delete an outbound webhook	Yes
POST	/webhooks/{id}/test	Queue a webhook.test delivery to this endpoint	Yes

Account

Your own audit trail and invoices.

Method	Path	What it does	Idempotency-Key
GET	/account/audit	Your own audit trail: sign-ins, security changes, purchases, portal actions and API writes, newest first (read scope)
GET	/account/invoices	Issued invoices for your account, newest first — the same list the billing page shows (read scope)
GET	/signup	For an assistant helping someone WITHOUT an account: plans, prices and a pre-filled sign-up link the person opens and completes themselves. Public (no token), creates and sends nothing
GET	/account/team	Members and their roles, and pending invites (read scope; changing the team is done in the portal)

Orders

Orders an assistant or script proposes for an owner or billing member to approve. Nothing is bought without that approval. Needs the billing scope.

Method	Path	What it does	Idempotency-Key
GET	/orders	Orders proposed on this account and their state (billing scope)
POST	/orders	Propose an order for the customer to approve (billing scope). Nothing is bought: owners and billing members are emailed and approve at approve_url	Yes
GET	/orders/{order}	One order: state, price, approve_url and, once approved, where it led (billing scope)
DELETE	/orders/{order}	Withdraw an order still awaiting approval (billing scope)	Yes

Related

• API reference
• Scopes and permissions for tokens and assistants
• MCP tool reference

Stuck? Contact support — replies within one business day.