SuperAPI AI Directory

done-done

General & other

View repository

Updated 10-17 16:42

About this skill

done-done

"No proof, no done." done-done captures proof from the real product, ties every screenshot, recording and log to the exact commit, and checks that cited proof is real, fresh and not reused. You run it for the user; they only say what should be proven.

Current guidance lives in the CLI

Do not follow command or config details from memory or from this file - installed copies go stale. Get the current source of truth from the CLI.

First run command -v done-done. If it prints a path, run done-done as shown below. If not, run npx -y done-done-axi wherever this file or done-done output says done-done - nothing to install.

  • done-done (no arguments, inside the repo) for this repo's proof setup and the next step
  • done-done --help for every command and flag, the journey format and the exact ## Checks block
  • done-done <command> --help for one command's usage and flags

Results are TOON: read result and follow the help lines through init, doctor, run, verify, check and receipt. Proof comes from driving the real product, not from a unit test run, and counts only from a committed tree: commit the done-done setup (config, capture script) before you capture. fail or inconclusive is not done: fix the cause and re-run, or tell the user exactly what is still unproven. A run waiting at its gate (awaiting_approval) is the user's call: show them its findings and captures, and run done-done respond only with their answer and reason. Never approve or skip on your own; overridden is not done either. A journey that spends runs only after the user approves spending: done-done run <journey> refuses it, so ask the user and use verify.

For a web page, done-done look <url> is the look step a journey calls: it captures the page at a 390x844 phone viewport with its page-check record and exits 1 on squashed, clipped or cut-off UI, naming each element. A hit fails the journey; never wrap it in || true or log the exit and go on. A journey in .done-done.yaml that calls it:

journeys:
  inbox-look:
    command: done-done look http://127.0.0.1:3000/inbox --out "$DONE_DONE_EVIDENCE_ROOT/inbox.png"

done-done starts a coding agent the user chose. When it says no agent is set, ask the user which agent (and model) to use and save their answer with the command it prints; never choose one for them.

done-done runs its own agent: done-done verify drives the journey, then a judge the user chose, never the agent that did the work (never you), looks at each capture. That is on by default; skip it with --no-judge only when the user chooses to. When it says no judge is set, ask the user which judge (and model) to use and save their answer with the command it prints; never choose one for them.

Request

$ARGUMENTS

If the request above is non-empty, the user invoked /done-done explicitly - fetch the current CLI guidance, then prove that. If it is empty, infer what to prove from the conversation.