Give your AI a big job.
Get proof back.
goalify turns your request into a goal: a clear finish line. It asks all its questions first. Then the AI works alone, for hours or even days. It stops when the proof is there, or at the limit you set.
Works best in Claude Code and Codex. Also installs in 70+ other AI tools.
claude plugin marketplace add Aboudjem/10xclaude plugin install goalify@10xnpx skills add Aboudjem/goalify44 seconds. Sound optional.
The AI said done. It was not.
Long AI jobs fail in two ways. The AI stops to ask you questions. Or it says "done" when it is not done. goalify fixes both before the work starts.
WITHOUT GOALIFY
- It stops at 3am to ask you a question
- The plan is lost when you clear the chat
- It uses a test your project does not have
- "Done!" only means it said so
- You find the problems the next morning
WITH GOALIFY
- All questions come first, one time
- The plan is saved in a file the AI reads again and again
- The tests come from your real project
- "Done" means it shows the tests passing
- One command runs the checks again and gives you a receipt
The thinking comes first. Then it just works.
goalify only prepares the work. The /goal command does the work. goalify makes sure the AI has nothing left to ask.
Homework
Can this work? It reads the docs, looks at your project and finds your real tests.
Questions, once
A few real choices, all together. After that, it does not ask again.
Brief + condition
A file with the full plan, and one short line for /goal that says what proof looks like.
/clear, paste, walk away
The work starts in a clean chat, with all its memory free for your job.
No interruptions, because there is nothing left to ask.
The research, your choices and the finish line are all ready before you type /clear. Then the AI works, checks, fixes, and repeats. In Claude Code, a second checker reads the work after every step. It lets the work stop only when the proof is there, or at the step limit you set.
See the plan before. Check the result after.
Run preview NEW
A page that shows the whole plan: the steps, the checks that prove it is done, the limits and your choices. It has a copy button for the /goal line. It also refuses to show a broken plan.
Proof receipt NEW
Runs the checks again. It shows you the commands first and asks one time. Then it writes a receipt: VERIFIED, NOT VERIFIED or UNVERIFIABLE. It can say no, even when the AI said done.
/goalify verifyReal checks, found for you NEW
Looks at your project files and your CI to find the real tests. It skips empty tests, so the finish line never uses a test that cannot pass.
part of every /goalifyStatus NEW
Lists every plan: open, blocked or done. For each open one, it shows the /goal line to continue. It changes nothing.
Express mode NEW
For medium jobs: no web research and no questions, but the same plan and the same safety checks. If the job is not clear, it stops and asks you to use the full mode.
/goalify --quick <task>Doctor NEW
Lists every copy of goalify on your computer. It warns you when an old copy is still loaded.
/goalify doctorWhere goalify fits
Different tools do different jobs. Here is where each one fits.
| Tool | What it is | How it differs from goalify |
|---|---|---|
| goalify | A skill that prepares the work for /goal: plan, finish line, preview, receipt. | It does not run the job itself. It uses the built-in /goal, so you install no extra loop. |
| supergoal | A Claude Code plugin. It explores, asks, plans in phases and gives /goal one command. It checks the result against the plan at the end. | The closest tool. goalify writes one full plan and a checked finish line, shows a preview page before the work, and writes a proof receipt after. |
| Ralph loops | A loop that gives the AI the same prompt again and again. | It usually stops on a word the AI writes, or after a number of rounds. With Claude Code's /goal, a second checker reads the work every step. |
| Spec Kit | GitHub's toolkit to write a spec, a plan and tasks. | Great for writing the spec. goalify is for long work with nobody watching, and for proving it is finished. |
Good to know
Does goalify do the job itself?
No. goalify never does the job itself. It writes the plan and the finish line. It runs your checks only when you ask (/goalify verify). The AI does the job in a new chat that you start with /goal.
Why not give /goal the plan file?
Because /goal needs a sentence, not a file path. The checker behind it cannot open files. So a path alone can never be checked. goalify puts the path inside a full sentence instead.
If the work stops, is it finished?
Not always. The checker can also stop the work if it thinks the goal cannot be reached. Run /goalify verify to run the checks again and get a receipt.
How much does it cost?
goalify is free (MIT license). It never guesses a price for your work. The step limit is your control.
Does it send my data anywhere?
No. No server, no extra packages, no network calls. The scripts use only standard Python, and the preview page loads nothing from the internet.
Does it work with Codex?
Yes. Codex has its own /goal, so goalify also gives you the Codex line. In Codex, the finish line works the same. Some of the work rules in the plan may not.
Give your AI a big job.
Get proof back.
claude plugin marketplace add Aboudjem/10xclaude plugin install goalify@10xThen type /goalify and describe the job. Needs Claude Code 2.1.139 or newer.