Real runs. Real numbers.

The Leadsets blog publishes real runs with real numbers, grouped into demos, signals, benchmarks, skills, email and building in public. The featured demo shows Claude finding 100 SaaS sales leaders from one prompt, including every tool call, what it cost and what the team would fix. To see the tools those runs use, read about Leadsets for AI agents.

Yes. The post "We compared 1,000 work emails across providers" covers the method, the mistakes and why catch-all disagreement matters more than the winner. Leadsets publishes its benchmark method and does not promise 100% accuracy, as explained on the Leadsets data page.

The blog post on hiring signals explains why hiring posts are a strong timing signal in B2B and how to pull them in one call. In Leadsets, scrape_jobs turns a careers page into open roles with titles and dates, and search_people then finds who runs sales there. See how the company and careers page scraper works.

No. The Leadsets blog treats them as two different jobs that often get sold as one: finding an address is not the same as checking it. Leadsets returns verified work emails from enrich_person and checks existing addresses with verify_email, using the waterfall described on the email verification page.

The post "A free Claude skill for B2B prospect research" covers what is in the skill, how to install it and how to fork it for your own ICP. The skill is part of a library of 50 free, MIT-licensed playbooks in the Leadsets skills library.