📇 LEAD GEN

How to build a Clay-style lead list with Jev

Clay-style lead tools charge a monthly subscription to find the right people at the right companies. Treg rebuilt the idea for AI agents and charges per lead instead, with Jev deciding which companies are worth looking up. It’s on GitHub. Here’s how to install it, what the licence lets you do, and a prompt that builds your first list.

Written September 28, 2026 · By Ben Broch
candidates screened
320
kept as relevant
96
start to finish
37.2s

From @jasonzhou1993’s public demo

Get Treg on GitHub ↗Let Claude Code / Codex build it ↘

The code: on GitHub, and it uses Jev

Jason Zhou’s team built Treg, a single key that gives an AI agent access to thousands of data tools: people and company search, work emails, phone numbers. The repo really uses Jev. It judges each company against your ideal customer from the free list data before any paid lookup runs, so you only pay to enrich the companies worth it.

3.7k ★stars on GitHub, September 28
Dailycommits: last one September 28
Source availablefree to self-host for your own team; not to resell as a service
Yesuses Jev to judge each company
Install Treg
# From the Treg README (I haven't run these myself)
curl -fsSL https://treg.to/install.sh | sh     # install the CLI
treg login                                     # GitHub, email code, or --token
treg catalog search "find work email for a person"
treg balance                                   # what that cost

# Or as a Claude Code plugin:
/plugin marketplace add superdesigndev/treg
/plugin install treg@treg

Repo: github.com/superdesigndev/treg · Hosted version: treg.to/people-search. New teams get $1.00 of free credit, according to the README.

01

What the demo did

An agent typed “Find head of growth at Series-A startup in SF.” Jev screened 320 candidates in 37.2 seconds and kept 96 as relevant, for a scan cost of about a cent. Then Treg found phone numbers and buying signals for the shortlist.

The pitch in the post: no more $600 subscriptions, just $0.0089 a lead, and 85% cheaper than Clay.

This is a product launch, and Treg sells the product. “#1 on people search bench” is their own claim, and the names in the demo table are sample data.

02

Judge first, pay second

Finding a person and verifying their email costs real money on every row. Asking Jev whether a company fits costs a tiny fraction of that. So the order matters: let Jev throw out the wrong companies from the free fields, and only pay to look up people at the ones that pass.

The same pattern with Jev

One call per company:

The agency question does the job the vendor question does on the prospecting guide: it removes the companies that look like a fit on paper but never buy. That guide finds people who are asking for what you sell. This one builds a list of people who fit. Most teams want both.

New to Jev? Start with the setup guide: key, first call, working response. Then come back.

Let an agent build it

Paste this at Claude Code or Codex once Treg is installed. It shows the price before each paid step and never contacts anyone. The thresholds are starting points I haven’t run yet.

Paste at Claude Code / Codex
Build me a lead list using Treg for the data and Jev for the judging. Treg is set up (the treg CLI or the Claude Code plugin); Jev runs through the Vercel AI Gateway.

1. Ask me who I sell to, in one plain sentence (for example: "B2B software companies with 50 to 200 staff that sell to sales teams"), and the role I want at each (for example: "Head of Growth or VP Marketing").
2. Use Treg to pull a first list of about 50 companies that roughly match on size and location. Show me the price before you run it.
3. For every company make ONE call: experimental_evaluate from the ai package, model typesafe-ai/jev. The state is the company's list fields (name, description, industry, headcount). Ask boolean questions together in that one call:
   - fit: "This company matches: <my one sentence>."
   - b2b: "This company sells to other businesses, not to consumers."
   - agency: "This company is an agency or consultancy, not a product company."
4. Keep a company only if fit >= 0.60, b2b >= 0.60 and agency < 0.50. Only now spend money: for the keepers, use Treg to find the person in my role and their work email, with the cheapest finder that only bills on a hit, and verify the email.
5. Run 3 Jev calls at a time; bursts get rate-limited. Write leads.csv (company, person, title, email, verified, fit), strongest fit first, and tell me the total spend by step.

Stop there. Never email, message or connect with anyone yourself. I review the list and reach out by hand.

The honest part

I haven’t run Treg myself yet. The numbers are from @jasonzhou1993’s launch post of September 23, 2026, and the install steps are copied from the README. I only publish code I’ve run, so there’s none here.

A list isn’t permission. Outreach to people in the EU and UK needs a lawful basis, and every platform has its own rules. Use the list to be relevant, not to blast.

Don’t have Jev running yet?

The setup guide is one key and a few lines of code, and you’ll have a working response in a few minutes.

Get started with Jev →

More field guides

All guides →
Jev · GuideHow to set up Jev and make your first API call →Step by step, about 5 minutesJev · GuideHow to automate SEO internal linking with Jev →The script, the prompt and the threshold ruleJev · GuideHow to find sales prospects ready to buy with Jev →Two scripts and the three filtering questionsJev · GuideHow to automate bookkeeping with Jev →Categorize every transaction, flag what needs a humanJev · GuideHow to filter Amazon products fast with Jev →The agent browser, the rules as questions, the raceJev · GuideHow to search Zillow by what the filters can’t see with Jev →699 listings judged in under 18 secondsJev · GuideHow to check legal citations and research state law with Jev →LegalJev (free) and Monid, side by sideJev · GuideHow to find buyers on LinkedIn with Jev →290 posts scored in 3.8s, 23 buyersJev · GuideHow to sort your inbox with Jev →500 emails in 7 seconds for a centJev · GuideHow to find viral TikToks with Jev →780 videos, 137 outliers, sorted by nicheJev · Directory200+ Jev demos, repos and docs by team →Everything worth watching or runningHiggsfield · GuideHow to swap the people in any video with Higgsfield Genjutsu →The clip, the photos, the prompt and the costChatGPT · GuideHow to build and submit your own ChatGPT plugin extension →The server, the ZIP and the review checklistClaude Code · GuideHow to make motion graphics with Claude Code →Six looping effects, one prompt eachClaude Code · GuideHow to security-check your vibe-coded app with Claude Code →One prompt, 20 checks, a fix listClaude Code · GuideThe UX checklist: make your app easier to use →The UX laws, a checklist per screen and one audit prompt