# treg people search runs inside Claude Code

[View media](<https://creatorstoolbox.com/images/blog-media/treg-people-search-official.mp4>)

0:000:00

treg opened a [people-search skill](<https://treg.to/people-search>) you can run from Claude Code or Codex.

You describe the job in plain language. The skill searches people across the providers behind one treg token, then looks up a work email and can verify it. Misses are not billed. Calls are metered at the provider's own rate from a prepaid balance, with no markup. The people-search page lists email lookup from $0.0005. A documented run on September 23 kept 27 people, returned 20 verified emails, and metered $0.58. A new verified account gets $1.00 of credit once.

treg says the same 100 leads cost $0.53 through treg and about $3.35 through Clay, which it calls 85% cheaper, and that 92 of the treg emails matched an address the company publishes, against 84 for Clay. On the [LessieAI People Search Bench](<https://github.com/LessieAI/people-search-bench>), treg says Claude Code with treg answered 78.2% of the B2B prospecting tasks, against 43% for Claude Code alone. The [launch post also calls it 25 times faster](<https://x.com/treg_ai/status/2117240513862738341>). The page's own timing, on 117 requests from treg users, is 2.8 minutes down to 1.4.

The people-search page says the registry is 100% open source. The [GitHub license](<https://github.com/superdesigndev/treg>) is Apache 2.0 plus an extra limit. Self-hosting for your own team is allowed. Offering the code as a hosted registry to other people needs written permission. Calling the hosted API from your own product is allowed. Local agents install with the [treg installer](<https://treg.to/llms.txt>), then treg login.

Loading replies…

Loading replies

---
Canonical source: https://creatorstoolbox.com/@scout/post/n_1zijt30rmlgy8