OpenSource
LinkedIn Recruiter

Paste a job post into your agent. It ranks candidates on their full social footprint and finds each one's personal email and mobile, for $0.19 a candidate. No seat, no InMail limit, open source.

Set up treg GitHub Repo
49 seconds · the candidate in the film is an example; every price and benchmark number is measured

Same candidate.
30x cheaper.

A Recruiter Lite seat rents you 30 InMails a month and no way to reach anyone outside LinkedIn. Your agent finds the same person and the contacts LinkedIn won't show you, and you pay per candidate.

LinkedIn Recruiter Lite
Maya RaoStaff Backend Engineer · London
🔒$170/month
30 InMails · no email · no phone
$5.67per candidate reached
Claude Code + treg
Maya RaoStaff Backend Engineer · London
✉maya.rao.dev@gmail.com✓
☎+44 7700 900123✓
$0.19per candidate, personal email + mobile included
Recruiter Lite: $170 ÷ 30 InMails. treg: email $3.18 + phone $1.39 + verify $0.03 over 24 candidates on a recorded run

Matched on her
full social footprint.

A LinkedIn title says what someone is called. The agent reads the job post, then checks each candidate where their work actually shows up, from career history to code, talks, papers and posts, and cites every source it used.

treg searches 14 sources per candidate
Who they are
LinkedIn search
profiles by title, skill, place
LinkedIn posts
what they write about
X posts
what they talk about
What they've built
GitHub
repos, stars, commits
Stack Overflow
answers in their stack
YouTube talks
conference talks
Google Scholar
papers
Whether they'd move
Job changes
tenure and moves
Company news
layoffs, funding, reorgs
Crunchbase
their employer's story
How to reach them
✉Personal email
not their work one
☎Mobile phone
cheapest first
Email verify
deliverable, won't bounce
Carrier check
live mobile, not a landline
Maya RaoStaff Backend Engineer · Ledgerly · London
example candidate
0match to the job
the job askswhat the agent found, and where
5+ years backendfrom the job
9 years, all backendBrightpay → Kestrel Bank → Ledgerly
✓ match
Built payments in Rustfrom the job
Wrote a payments ledger in Rustrust-ledger ★ 1.2k on GitHub · a talk on moving £2B a day to it
✓ match
Based in Londonfrom the job
Lives in LondonShoreditch · 25 min to the office
✓ match
Fully vested in 2 monthswhat you can't search for
Joined Ledgerly Dec 20224-year vest ends Dec 2026 · the usual moment to move
✓ timing
evidence from 7 sources · about $0.02 a candidate on a recorded run

Her personal email.
Her mobile.

treg asks every vendor that finds personal emails or mobiles, cheapest first, and stops at the first answer. A miss costs nothing. Emails are checked deliverable; numbers are checked live with the carrier.

✉Personal emailnot her work one
Fiber$0.04waiting
Aviato$0.08waiting
LimaData$0.10waiting
ContactOut$0.25waiting
Sixtyfour$0.48waiting
✓ verified personal emailmaya.rao.dev@gmail.com
☎Phone numberher mobile
BlitzAPI$0.0005waiting
QuickEnrich$0.0048waiting
AI Ark$0.026waiting
Tomba$0.044waiting
Wiza$0.13waiting
✓ live mobile+44 7700 900123
example lookup · vendor order and prices are the catalog's

The best personal email finder
is all of them, stacked.

We sent the same 30 real candidates, by LinkedIn URL, to every vendor and verified every answer. No single vendor got past 77%. Stacked, treg found 80% for a third of the price per email.

Verified personal email found30 candidates · cost per valid email
EnrichLayer10%$0.13
LeadMagic40%$0.06
Fiber43%$0.04
LimaData50%$0.11
Aviato53%$0.09
ContactOut best alone77%$0.28
treg all stacked80%$0.10
Live mobile found30 candidates · checked with the carrier · cost per live mobile
AI Ark23%$0.049
LeadMagic30%$0.167
Tomba40%$0.037
Prospeo n=1553%$0.276
Wiza best alone53%$0.195
treg all stacked57%$0.21
benchmarks run 2026-10-17 · same 30 people sent to every vendor · emails verified deliverable, mobiles checked live (live is not proof it's the right person) · cost = what the vendor billed ÷ valid results

Every people-data vendor.
One key.

Your agent asks for a job, not a vendor. treg tries the vendors that can do it, cheapest that answers first, and a miss on a pay-per-hit vendor costs nothing. Bring a key you already pay for and treg uses it first, never marked up.

Find people
CrustdataPeople Data LabsCoresignalExaApolloAviatoIcypeasLeadsforgeSumbleDiffbotParallelCrunchbaseHarvestAPICompanyEnrichMoltsetsEnrichLayerLushaContactOutLeadMagicProspeoWizaAI ArkBlitzAPIQuickEnrichFindymailHunterDropleadsGetLeadsOcean.ioFiberLimaDataSixtyfour
Mobile phone
BlitzAPIQuickEnrichAI ArkTombaWizaProspeoLeadMagicLeadsforgeLushaContactOutDatagmaDropleadsFindymailGetLeadsOcean.ioAviatoLimaDataSixtyfour
Personal email
FiberAviatoLimaDataContactOutSixtyfour
Verify
MillionVerifierZeroBounceHLR LookupTrestle

What a seat buys you

Prices are LinkedIn's list prices as published by third parties; check LinkedIn's own page for your region.

Recruiter LiteSales Navtreg
Price$170/mo$99.99/mopay per candidate
Reach30 InMails50 InMailsno limit
Personal email✕✕✓
Mobile✕✕✓
Per person reached$5.67$2.00$0.19

How a run goes

The recruit skill turns your agent into the sourcer. It shows you the criteria before it spends anything, and every stage is counted in a run file you can check.

  1. Read the jobFrom the LinkedIn post URL, or the text you paste.
  2. Agree the criteriaTitle variants, must-have skills, seniority, location. Shown to you in five lines first.
  3. SearchOne routed people search per title variant, across the people-data vendors.
  4. Filter and rankDrops the wrong location, seniority or job family, with a reason for each. Ranks the rest with one line on why.
  5. Check the evidenceCited facts from career history, code, talks and posts move a candidate up or down.
  6. Find the contactsPersonal email and mobile for the shortlist, cheapest vendor first, each one verified.
# in Claude Code, Codex or any agent with treg set up
› find candidates for linkedin.com/jobs/view/4021877351

Give your agent
a recruiter seat.

$ set up treg — treg.to/llms.txt
$ npx skills add superdesigndev/treg --skill recruit
Get started Open source on GitHub