One AI task, one CLI script, runs locally.
You want a working starter to validate AI fits your problem, before committing to a hosted agent.
- Working CLI script (Python or Node) doing one specific task: categorize, extract, transform, summarize, or translate
- Custom prompt tuned to your input/output spec
- 5+ sample inputs run with their outputs included (proof it works)
- 1-page README: setup, how to run, how to change the prompt
- Delivered as zip + pushed to your GitHub
- 1 written Q&A round (up to 3 questions, 1 business day response)
- Hosting or deployment (runs on your machine, see P02 for hosted)
- Integrations with other systems
- More than one task per engagement
- Iterations after delivery (Q&A only, then closed)