Run & output

Hit Run and Typelets executes your entrypoint in the workspace sandbox, streaming stdout and stderr back to the browser as they are produced.

Choosing what runs

You can run a specific file, or let Typelets pick the entrypoint:

• If you run a specific file, it executes that file (as long as it exists in the workspace).
• Otherwise Typelets auto-detects an entrypoint, preferring a main.* file, then common names like index, app, program, or solution, then a lone runnable file. If it cannot decide, it asks you to add an entrypoint.

Run handles single-file programs in the supported languages - see Languages & runtimes. Multi-file projects, app servers, and build pipelines run from the terminal or via a preview startup command instead.

Live output

Output streams back as it happens, not after the program exits, so you see progress in real time. Output is bounded to keep a runaway program from flooding the browser:

• Time limit: a run is killed after ~20 seconds.
• Output limit: ~1 MiB total, rate-capped at ~256 KiB/second. Beyond that the stream is marked truncated and the run settles.

Who can run

Running requires an editing role (not viewers) and a verified email address. Viewers can watch output but not trigger runs.

Test runs (interview mode)

Interview workspaces also have a test runner that executes the problem’s test cases - either stdin/expected-output comparisons or framework commands - and reports pass/fail. Candidates see hidden test cases only as pass/fail, never their expected output. See Running an interview.