Skip to content

API command index ​

Flowstride's API adapter uses Undici. Relative endpoints use BASE_URL from .env or configured baseUrl. API and browser commands can share the same .flow file and extracted/generated data; see API and UI in one flow.

TaskReference
Readflow.get
Createflow.post
Replace/updateflow.put, flow.patch
Deleteflow.delete
GraphQLflow.graphql
Retain response dataflow.extract
Check resultsflow.expect
Send JSON, form data, text, a multipart upload or binaryflow.api.request
Isolate API users and cookiesNamed API contexts
Validate a JSON response shapeflow.api.expectSchema
Wait for an API job or state transitionflow.api.poll
Capture or mock a page requestBrowser network
flow
flow.post "/api/orders" with reqHeader "Authorization" "Bearer {{env.API_TOKEN}}" with reqBody
"""
{
  "sku": "DOCS-001",
  "quantity": 1
}
"""
flow.expect status "201";
flow.expect resBody "id" not.empty;
flow.extract resBody "id" as "orderId";

Use strict camelCase for reqBody, reqHeader, resBody, resHeader, and responseTime. Triple-quoted docstrings keep JSON readable. Assert a response before relying on data extracted from it.

Browser network routes affect browser traffic, not these direct API calls. API calls and a browser page are separate clients. Do not assume every cookie/token from one is automatically active in the other; use an explicit session workflow and verify the resulting application behavior.

For HTTP methods, chained comparisons, GraphQL variables/mutations and explicit API-cookie browser login, read API testing. For a complete local application and deliberate failing check, use the included examples.

Flowstride is MIT licensed. Flowstride Cloud offers free and paid workspace plans.