Worksheet validation
The Python checker has 12 offline checks covering missing values, invalid counts and failure handling. The optional checker above makes this workflow usable without a terminal.
Inspect checker and testsOPTIONAL · FOR DEVELOPERS RUNNING THEIR OWN TESTS
Supporting resources.
The checker below is useful if you already have manually recorded test results. It finds missing details and summarizes those records; it does not run a browser, collect timings or make tasks faster.
No CSV is needed to use BrowseSprint’s guide. Start with practical changes you can try.
AVAILABLE NOW · FREE · NO ACCOUNT
Check your browser-task records.
Paste a completed worksheet. See missing fields, separate failures from successes, and inspect timings for each recorded context.
Processed in this tab. CSV contents are not uploaded or saved. Use public-page examples; leave out credentials and private account data. Up to 128 KB / 300 rows.
WHAT WE HAVE ACTUALLY TESTED
Inspectable code. Explicit limits.
These small controls demonstrate result checking. They are not evidence that an unreleased extension makes browsing faster.
A normal return can fail the task.
Two AutoGen FunctionTool calls returned HTTP 200. The correct page title passed; an intentionally wrong title failed the separate check.
Code, dependencies and actual JSONCheck extracted results too.
The Crawlee CheerioCrawler example repeats the title-check control with public HTML. It uses direct HTTP, without a browser or model.
Inspect executed examplesPrepared with AI assistance. Source links are our own resources; they do not imply endorsement by the projects named above. Controlled Codex timing comparisons have since been run. Read the measured results, negative cases and limits. They do not establish a general speedup or test a released extension.