I built CLISHOP: an open-source CLI for AI agent shopping

I’ve been building CLISHOP, an open-source CLI for AI agent shopping.

The core idea is simple:

What if an AI agent had one interface to search, compare, and complete a purchase flow across connected stores — with guardrails like approval steps, spend limits, and confirmation gates?

Website: https://clishop.ai/

GitHub: https://github.com/DavooxBv2/CLISHOP

Why I built it

A lot of agent tooling today focuses on reasoning, browsing, or calling APIs.

But commerce is still awkward.

If you want an agent to actually help with buying something, you quickly run into messy problems:

• every store has a different flow
• trust matters a lot more when money is involved
• you need controls, not just autonomy
• even testing the concept is harder than it sounds

So instead of building another general AI wrapper, I wanted to explore a more specific question:

What would shopping infrastructure for AI agents actually look like?

What CLISHOP is

CLISHOP is a CLI that aims to give agents a unified interface for commerce tasks.

That includes things like:

• searching products
• comparing options
• running purchase flows
• working across connected stores
• adding safety layers like spend caps and confirmation gates

The idea is not “let AI spend money with zero oversight.”

It’s more like:

give agents a structured commerce interface with explicit controls.

What the CLI experience looks like

The experience is designed to feel simple from the command line.

Example flow:

bash
npm install -g clishop
clishop search "wireless headphones"
clishop buy <product_id> -y