From OpenAPI spec to tested requests
Import an OpenAPI spec and get a collection of runnable requests in the Request Workbench. Configure OpenAPI Sync to reconcile changes from your spec and review what changed in the collection. Add flow assertions to check the API itself.
Upload a spec, get a working collection
Import an OpenAPI spec and every operation becomes a saved request in a collection — organized into folders, searchable by operation. No retyping paths, no copying schemas by hand.
Collections hold everything side by side: imported APIs next to requests you wrote yourself, with search across all of them. Review the imported URL, auth, parameters, and body before sending a request to your test environment. The Request Workbench guide covers those controls.
- "Import OpenAPI": spec upload → collection of requests
- Folders for structure, search for everything else
- Same import available from the CLI for scripting
Run any operation in the Request Workbench
Every saved request opens in a multi-tab editor with Params, Headers, Auth, and Body. Auth helpers cover the usual patterns, including API-key placement options, so you spend your time on the request — not on remembering where the key goes.
Requests execute server-side, and the response body and headers land right next to the request — a wrong status code or an unexpected content type is visible immediately. Edit, send, read, repeat — across as many open tabs as the investigation needs, with every request saved for the next session.
Track changes between the spec and your collection
Enable URL-based sync per collection and a background process reconciles spec changes every five minutes. Want an on-demand sync? Trigger run-now and read the result. Because the configuration is per collection, a stable internal API can sit untouched while the fast-moving one reconciles continuously.
Every sync is recorded: the drift and sync history shows what was added, changed, or removed, and operation search takes you straight to the endpoint you're worried about. API drift stops being something you discover from a changed collection. Sync can overwrite generated-request edits when an upstream operation changes; it does not rewrite existing flows. Review the OpenAPI import and sync guide for configuration and history.
- Per-collection sync configuration
- Run-now for on-demand reconciliation
- Drift and sync history, plus operation search
Scriptable imports from the command line
The same import runs from the terminal:
echopoint collections import --file openapi.yaml.
Wire it into the pipeline that builds your API, and collections track the spec
your team actually ships — not the one from last quarter.
The CLI is public on GitHub and installs with one checksum-verified command. See the CI/CD page for the full run-in-CI story.
echopoint collections import --file openapi.yamlWhat happens when your orders API changes?
Suppose your spec defines POST /orders and GET /orders/{id}. Importing it gives you saved requests to inspect and run; it does not prove the live API behaves correctly.
Import a snapshot
Upload the spec or use collections import --file. Open a generated request and supply the base URL, auth, and test data it needs.
Connect the spec URL
Configure OpenAPI Sync for the collection when you want it to follow future spec changes. A one-time file import does not establish continuous sync by itself.
Review collection drift
If the spec adds an operation or changes a request definition, inspect the sync history. This compares the spec with saved requests, rather than testing live responses against every schema.
Test the behavior
Build a flow that creates an order, extracts its ID, and reads it back. Add status and body assertions, then run that flow against your test environment.
Your agent can use the CLI or MCP tools to work with collections and flows. See the request-chaining example, configure response assertions, and run API tests in CI.
Import, organize, execute, sync
Import
Upload a spec in the app, or run collections import --file from the CLI.
Organize
Folders keep collections tidy; search finds any operation across all of them.
Execute
Params, Headers, Auth, Body — server-side execution in the Request Workbench.
Sync
Enable per-collection sync, trigger run-now, and review the drift history.
Keep requests true to the spec.
Join the waitlist for access to OpenAPI import and sync. While you wait, review the docs and choose a spec to test. Free while in beta, no credit card.
