Build an AI agent that runs one operation of your business.
Pick a template. Copy one instruction into Claude Code, Codex or Cursor. Your coding agent builds the agent in your own repository. Free, no account.
How it works
- Choose a template below and read what a human still does and what it needs.
- Press Copy, and paste the instruction into your own coding agent.
- Your coding agent builds the agent, in your own repository, on your own credentials.
Paste it into a Claude Code conversation, or save it under a project's own skills folder and reference it by name.
Paste it into a Codex conversation, or add it to your project's AGENTS.md so Codex reads it automatically.
Paste it into Cursor's chat, or save it as a rules file under .cursor/ and reference it.
Paste it into any coding agent that can create files and run shell commands - the instruction names what it needs from the agent, not which one.
Your coding agent runs on your own subscription or API key. Provek runs nothing for you and sees nothing you build.
Templates (6)
customer support: inbound email triage and reply drafting
- Human still does:
- sending any reply; issuing refunds; anything with legal or safety content
- Needs:
- an inbox reachable by IMAP or a provider API (e.g. Gmail API); an LLM API key
- Dry run · 2026-09-05 · claude-code · scaffold produced; 6 of 6 template tests passed
lead generation: sourcing, enriching, and qualifying leads against a stated ideal customer profile, then drafting first-touch outreach
- Human still does:
- approving the ideal customer profile; sending any outreach message; deciding who gets contacted
- Needs:
- a source of company/contact data (a CSV export or a data provider the user already has rights to use); an LLM API key
- Dry run · 2026-09-05 · claude-code · scaffold produced; 6 of 6 template tests passed
e-commerce back office: performance summaries, listing hygiene, inventory/order alerts, and price-change proposals, all behind a human approval gate
- Human still does:
- approving every staged change (price move, restock, listing edit, or promotion draft) before it is applied to the live store; nothing here has a live write credential except the one approval step
- Needs:
- read access to the store's catalog/order/inventory data (an export, a database read replica, or the platform's read-only API); an LLM API key
- Dry run · 2026-09-05 · claude-code · scaffold produced; 6 of 6 template tests passed
market/competitor research: a recurring brief produced by asking several models the same questions independently, having them review each other's anonymized answers, and combining the results into one sourced document
- Human still does:
- choosing the research questions each run; reading the brief before it is shared or acted on; deciding what to do about anything it finds
- Needs:
- API access to at least two different LLM providers or model families (the cross-review needs genuinely independent models, not the same model called twice); a way to fetch or paste source pages (a search API, or URLs the user supplies)
- Dry run · 2026-09-05 · claude-code · scaffold produced; 6 of 6 template tests passed
content production: drafting an article from a brief, checking every factual claim against the sources it cites, and queuing the result for a human to publish
- Human still does:
- approving the brief and its angle; publishing anything the program produces; deciding what to do with any claim the fact-check step could not confirm
- Needs:
- a content brief naming the topic, audience and allowed sources; a publishing target the human controls (CMS, static site, newsletter tool) - this template writes files, never publishes to it directly; an LLM API key
- Dry run · 2026-09-05 · claude-code · scaffold produced; 6 of 6 template tests passed
finance back office: invoice intake, categorization against a chart of accounts, and a reconciliation report comparing books to a bank/card statement export - it never moves money
- Human still does:
- reviewing and posting every categorized transaction to the actual books; resolving every reconciliation gap; any real payment, transfer, or payroll action - none of which this template performs
- Needs:
- an export of invoices/receipts (PDF or CSV) and a bank/card statement export (CSV); the chart of accounts (category list) the user already uses; an LLM API key
- Dry run · 2026-09-05 · claude-code · scaffold produced; 6 of 6 template tests passed
What follows
- 01 — An agent, in your repository
- Built by your own coding agent, from the instruction you copied.
- 02 — One business operation it runs without you
- Not a chat assistant: the operation the template names, with the human points it names too.
- 03 — A passport, if you request one and the repository is public
- An independent measurement of how much runs without a human, including what could not be measured.
- 04 — An Order link in the registry
- Once a passport for it declares where customers order from you.