Running, logs and reports

Run a flow with F5, read what every step did in the log, and write HTML reports of a flow or of a whole project.

Running a flow

Run Flow (F5) runs the open flow, top to bottom; the button becomes Stop while it runs. Each step's marker turns green or red as it finishes.

Run time limit

A flow that waits for something that never comes (a job that never finishes, a loop that never ends) would run until you stop it. In CI nobody is there to press Stop, so give the flow a limit:

  1. Open Flow Settings (the gear in the title bar).
  2. Under Running, set Run time limit (ms), for example 300000 for 5 minutes.
  3. Save the flow. The limit travels with it, into exported scripts too.

The limit counts the whole run: every step and every data row. When it passes, the run stops and is reported as failed, with Run time limit reached in the log and the reports. "Always run" steps and later rows don't run after it. A code block is cut off as well, even while (true) {}; code that never gives control back after an await ends the runner 5 seconds later, with the same message. Empty or 0 means no limit, the default.

Example: 23_run_time_limit polls a job that never finishes; the 15 s limit ends the run (it fails on purpose).

Exported scripts and the suite take --run-timeout <ms>, which overrides the flow's limit (--run-timeout 0 removes it). See Export and CI.

The Timeout (ms) next to it is per request. Use both: the request timeout catches one slow call, the run time limit catches a flow that never ends.

The play icon at the right end of Run Flow (or ↓ on the button) opens more ways to run:

The run menu: Run flow, Run flow with report, Run project and Run project with report
The run menu.
ItemDoes
Run flowThe same as clicking the button.
Run flow with reportAlso writes an HTML report of the run.
Run projectRuns every saved flow of the open flow's project.
Run project with reportThe same, with one report for all of them.

Reading the log

Each run adds a block to Logs & Responses:

The log of the tutorial flow: six passed steps with their requests and checks
A run's log.

Bodies are shown with viewers: JSON and XML as trees, CSV as a table, images and PDFs as previews, nothing cut. Copy icons copy a body or the headers; the copy icon at the end of a request line shows it as a cURL (or other) command, with the values it really had. Copy from the log.

The bin in the log's header clears it; the arrows at its side jump to the top or the bottom.

Masked values

Secrets, variables whose names look sensitive (password, token, secret, apikey, auth…) and credential headers (Authorization, Cookie, X-Api-Key…) show as •••••• in the log, the console output and reports. The run itself uses the real values.

Reports

Run flow with report writes report.html, the same report exported flows write with --report, into the run's output folder. The end of the run in the log gets a Report line, whose links pulse for three seconds when it appears: Open shows it in your browser and Show in folder shows the file. It is a single file you can send to anyone: every step, check and response, collapsible, with its own Dark / Light switch.

The HTML report of the tutorial flow: passed, 6 of 6 steps, 19 of 19 checks, and each step with its request and checks
The report of the tutorial flow, in both of its themes.

Running a project

Run project runs every saved flow of the open flow's project, one after another in name order, each in its own process with its own selected environment. A failed flow doesn't stop the others.

Flows without a project can't be run this way. To run a project in CI, export it.

Where run files go

Downloaded and saved files, and reports, go to <output folder>/<flow>/<date and time>. The output folder is set in Settings; the last 20 runs of each flow are kept.