Stacklytics. How long Flisk takes to bring each stack up.
Flisk's own CI scaffolds and starts every language, framework and resource type it supports on a fresh GitHub runner every night, and publishes the times here. Try one with npx @flisk-sh/flisk demo.
Not a runtime benchmark. A cold start here is the whole first flisk up on a fresh CI runner: pulling and building the image, installing dependencies and starting the container until the route answers. That is mostly Docker and package-install time, so it says how quickly Flisk gets a stack running, not which language or framework runs faster. Each number is one CI run: the same Bun stack took 47 s in one run and 27 s in the next, so differences of a few seconds are noise.
Start time
Per stack
Scaffold, cold and warm start for every stack in one view. Each bar is scaled to its column's slowest stack.
Cold vs warm
Each dot is one stack: first start across, second start (images cached) up. The further below the line, the more a warm start saves.
Warm-start saving
How much faster the second flisk up is than the first. Biggest saving on top.
Scaffold
flisk resource, from the command to the files on disk. Median of the leg's scaffolds.
Hot reload
Edit a file until the new response is served.
Trend
Median per kind of stack, one point per nightly run.
Coverage
Stacks Flisk supports, by kind: measured, ran without a timing, or not in CI yet.