MCP for Claude, Cursor & agents

800M+ verified
contacts for
your AI.

Claude, Cursor, and API access to people search, email & phone enrichment, and LinkedIn lookups — sub-200ms. Agents pay per call via x402. Humans start with 2,000 free matches.

Claude · AgenticData MCP
Connected
Find VP Sales at fintech companies in NYC. Return work emails and LinkedIn URLs.

search_people · 142ms

{
  "count": 12,
  "people": [
    {
      "name": "Sarah Chen",
      "title": "VP Sales",
      "company": "NovaPay",
      "work_email": "sarah@novapay.io",
      "linkedin": "linkedin.com/in/sarahchen",
      "city": "New York",
      "seniority": "vp"
    }
  ]
}
Message Claude…
// social proof

Built for teams shipping with AI

From the founder →
Early beta user
AI-native founder

I pointed Claude at the MCP endpoint and had a full TAM list in one conversation — no CSV export, no dashboard tab-switching.

Growth lead
B2B SaaS

Sub-200ms enrichment inside Cursor changed our outbound workflow. The agent enriches while it drafts.

Agency operator
Lead gen

Unlimited under fair use beats per-credit anxiety. We run batch enrich for client lists without watching the meter.

// field depth

30+ fields per person. Rich company context.

800M+ people and companies in an owned database — structured for agents, not spreadsheet dumps.

People · 30+ fields

Contact, location, job, profile

  • first_name · last_name
  • work_email · personal_email
  • mobile_phone · direct_phone
  • linkedin_url · headline
  • job_title · seniority
  • city · state · country
  • skills · work_history
  • email_verified · match_confidence

Companies · firmographics

Domain, industry, size, HQ

  • company_name · domain
  • industry · employee_count
  • revenue · funding_stage
  • sic_code · naics_code
  • hq_city · hq_country
  • linkedin_company_url
  • company_description
  • technologies · keywords

Processed for AI

Typed JSON responses

Every field is structured and schema-stable. Your agent reads records directly — no HTML parsing.

Freshness signals

Verification flags and match confidence on every contact so agents know what to trust.

MCP-native tools

enrich() and search() ship as first-class MCP tools. No export step between lookup and reasoning.

// setup

Connect your favourite AI

Hosted MCP — paste a URL and your API key. No npm install, no local CLI.

Remote MCP server URL
https://www.agenticdata.host/api/mcp
  1. 1Open claude.ai → Settings → Connectors (or Customize in Claude Desktop).
  2. 2Click + Add custom connector.
  3. 3Name it agenticdata and paste the Remote MCP server URL below.
  4. 4Press Connect, sign in, and allow access.
  5. 5Ask Claude to enrich an email or search for people by title and company.
$ enrich.sh
200ms · 1 credit
$ curl https://api.agenticdata.host/v1/enrich \
    -H "Authorization: Bearer $DATAAPI_KEY" \
    -d '{"email":"elon@spacex.com"}'

{
  "name": "Elon Musk",
  "title": "CEO",
  "company": "SpaceX",
  "linkedin": "linkedin.com/in/elonmusk",
  "mobile": "+1-555-…",
  "match_confidence": 0.97
}
§01 — Ledger

Seven numbers. That's the pitch.

Every claim below is a hard number. No adjectives. No “unleash.” No “supercharge.”

REC_001
Latency

Sub-200ms median. Every request.

We query an owned Postgres cluster. No scrapers. No queues. No wait-and-poll callbacks.

REC_002
Price

Agents: $0.002–$0.50/call. Humans: $1/2k.

x402 micropayments on Base for autonomous agents. API keys and subscriptions for teams. Misses are free on subscription plans.

REC_003
Finders

People, LinkedIn, phone, and email.

Reverse lookup by email, phone, or LinkedIn URL. Find work emails from name + domain. Search by title, company, and location.

REC_004
Records

800M+ verified people and companies.

Owned, licensed, refreshed weekly. 30+ fields per person including work email, mobile, LinkedIn.

REC_005
Protocol

MCP + x402. Payment is auth.

Agents use llms.txt, AgentCash, or MCP with a Base wallet. Humans paste an API key into Claude, Cursor, or any HTTP client.

REC_006
Scale

Batch 10,000 records in one request.

Synchronous batch endpoint. No webhook setup. Rate limits scale with plan tier.

REC_007
Fields

30+ fields per person.

Work email, personal email, mobile, LinkedIn, title, seniority, company, skills, work history.

§02 — Diff

How we differ from the incumbents.

They scrape while you wait. They charge you when they find nothing. We do neither.

FeatureAgenticDataLive-scrape toolLegacy data vendor
People / LinkedIn / phone / email finderYesPartialPartial
Agent x402 micropaymentsYesNoNo
Cost per 1,000 matches$0.50$1.00 – $2.00$1.50 – $3.00
Median latency<200ms5–60s20–90min
Charge on no-matchNoYesYes
LinkedIn URLIncludedLimitedAdd-on fee
MCP endpointNativeNoneNone
Owned databaseYesMixedYes
§03 — Why we exist

Your data
was never
theirs.

Incumbents rent you access to data they scraped on someone else's terms. We built an owned database and handed it to agents through MCP — you pay for compute, not permission.

Read the founder story →
OTTO — the AgenticData liberation droid
OTTO_v1 · open to the owner
§04 — FAQ

Common questions.

Q_01

What can I find?

People (by email, phone, or LinkedIn URL), work emails (name + domain), phone numbers, and LinkedIn profiles — plus company search and list building across 800M+ records.

Q_02

What does it cost?

Agents pay per call via x402 USDC on Base ($0.002–$0.50). Humans use API keys: $0.0005 per match ($1/2,000), misses free, or unlimited plans from $29/month.

Q_03

How fast is it?

Under 200ms median. We query an owned database. Scrape-based tools take 5 seconds to 2 hours because they scrape while you wait.

Q_04

Can I send my agent here without an API key?

Yes. Point your agent at agenticdata.host/llms.txt, fund a Base wallet with AgentCash, and pay per call in USDC. See /agents for the full setup.

Q_05

Why cheaper than incumbents?

They scrape LinkedIn or Apollo live — each lookup costs them a scrape. We own the database. Volume costs us nothing, so we hand you the savings.

§05 — Ship

800M+ contacts.
Start with 2,000 free.

Connect MCP in Claude or Cursor. No credit card required.