Build a 'Buy It For Life' AI Shopping Assistant with Claude

Start a new project in Claude to house your shopping assistant. In the project's instructions, define your core purchasing principles. This text will act as a persistent filter for all your searches. Include criteria such as product longevity, repairability, brand history, and clear return policies. Instruct the AI to be skeptical of trendy, heavily advertised brands and to watch for signs of drop shipping.

In the same project instructions, provide a list of vendors and stores you already trust. This gives the AI a starting point for its research, focusing it on a pre-vetted list of sellers before searching the wider web. For example, you could paste a list of curated retailers like Boston General Store or Manufactum.

Define a standard output format in your project instructions so you can easily compare recommendations. Specify the exact fields to return for each product: name, photo, price, materials, care instructions, a purchase link, and a summary of the brand's history. This structure makes tradeoffs between factors like price and materials clear.

After setting up the project instructions, ask for items with simple prompts. Because the detailed criteria are already saved in the project, your request can be brief. Claude will apply your philosophy and formatting rules to its search and return a list of vetted recommendations.

Use the same project to vet unfamiliar brands. Ask Claude to research a specific brand against your purchasing philosophy. It will return an analysis, flagging red flags such as a recent acquisition followed by poor reviews, negative employee feedback, or a heavy reliance on influencer marketing. Your judgment is needed to interpret these signals.