Build a lead finder that ranks prospects and checks emails

A developer made a Claude Code agent that discovers B2B prospects, scores them, and verifies email addresses for the strongest matches.

Done with Claude Codeยท Advanced

What they did
They spent a weekend putting the workflow together in Claude Code. The agent uses Tavily to locate actual business leads, then relies on Claude Code to judge which ones look like the best fit. Prospeo is only used on the top names, where it adds verified email addresses, and the design tries to avoid wasting API credits.
How it went
In the finished flow, the agent can uncover leads, order them by purchase likelihood, and enrich the highest-ranked ones with verified emails. The public write-up did not include performance totals, and the detailed build instructions were locked behind a paywall.
Tips
Keeping the research step separate from enrichment helps conserve free-tier credits. Before any outreach or export, have the agent stop and wait for your approval.

Try it yourself with Claude Code

Set up a lead-scouting agent in [project folder] with Claude Code. Have it search for businesses in [industry or region] that match [my ideal customer description], then assign each one a score from 1 to 10 based on fit. Use Tavily to research candidates, keep the search narrow enough to limit credit use, and for the top [number] entries, call [email lookup service] with [API key] to retrieve verified contact emails. Put everything into a CSV sorted by score, and before any sending, exporting, deleting, or submitting, pause and ask for my approval. Start with a test run on [number] companies and show the results first.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on medium.com โ†—

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.