Try it: your first build

Follow the first build step by step beside a terminal that replays a real run.

The same five steps as Your first build, with a terminal beside them. Type each command, or press Run this step, and the terminal plays back what that command printed in a real run, with long waits cut short. It is a replay: nothing you type here runs anywhere. To run the commands for real, install qq on your own machine.

  1. qq fetch clones a repo and runs qq sync in it, which downloads the repo's pinned toolchains and checks each against its digest. Make a folder for it first, and go into the clone after:

    mkdir -p ~/qq-try && cd ~/qq-try

    qq fetch https://github.com/quirq-ai/innernet

    cd innernet

    Your first build: innernetreplay
    # Replay of a real run. Type a command from the steps, or press Run on a step. Type help for more.
    ~ $ 
    Recorded 2026-10-21 on Linux. That machine could not reach ghcr.io, so it cloned with git and passed Node 24 in.
  2. A repo is a list of targets, each with a kind such as node-app that tells qq how to build and test it. Ask qqsync rather than reading infra/repo.toml yourself. innernet has one target, app.

    qqsync show

  3. qq test builds every target, then tests it, one line per step. Read each line as result, target, step.

    qq test

  4. Everything a run produces lands in .qq/out: results.json for scripts and agents, one log per step in logs/, and JUnit XML in junit/.

    ls .qq/out .qq/out/logs

  5. qq run runs a command with the repo's pinned toolchains first on PATH. innernet starts at http://127.0.0.1:3470 on your machine. The replay's first line, not synced here, is there only because the recording machine could not run qq sync; after qq fetch you will not see it.

    qq run "pnpm start"