Skip to main content
Glama
DataForB2B MCP Server

DataForB2B MCP Server

Official
by DataForB2B

DataForB2B MCP Server

An MCP server for the DataForB2B API — search and enrich companies and people (professional profiles), job postings, and social posts, directly from Cursor, Claude, or any other MCP-compatible client.

Tools

Tool

Description

search_people

Search professional profiles with structured filters (title, company, location, skills, funding stage, ...).

search_company

Search companies with structured filters (industry, size, country, funding, hiring, ...).

search_jobs

Search job postings on LinkedIn / Indeed — hiring signals.

search_posts

Search social posts on LinkedIn / Twitter / Reddit by keyword, or fetch a single post's engagers as leads.

enrich_profile

Enrich a single profile (full profile data, work email, personal email, GitHub).

enrich_company

Enrich a single company (description, headquarters, employee count, funding, offices).

Full parameter reference: docs.dataforb2b.ai/api-reference.

Related MCP server: apollo-io-mcp-server

Getting an API key

Sign up at dataforb2b.ai and grab your API key from the dashboard.

Install

Requires Python 3.10+. The recommended way to run it is with uvx, which needs no separate install step and always runs the latest commit on main:

uvx --from git+https://github.com/DataForB2B/dataforb2b-mcp dataforb2b-mcp

Or clone and install it with pip:

git clone https://github.com/DataForB2B/dataforb2b-mcp.git
pip install ./dataforb2b-mcp

Once published on PyPI, this collapses to plain uvx dataforb2b-mcp / pip install dataforb2b-mcp (the config examples below use the git form, which works right now).

Configure

Cursor

Add to .cursor/mcp.json (project) or ~/.cursor/mcp.json (global):

{
  "mcpServers": {
    "dataforb2b": {
      "command": "uvx",
      "args": ["--from", "git+https://github.com/DataForB2B/dataforb2b-mcp", "dataforb2b-mcp"],
      "env": {
        "DATAFORB2B_API_KEY": "your-api-key"
      }
    }
  }
}

Claude Desktop / Claude Code

Add to claude_desktop_config.json (Desktop) or run claude mcp add (Claude Code), using the same shape:

{
  "mcpServers": {
    "dataforb2b": {
      "command": "uvx",
      "args": ["--from", "git+https://github.com/DataForB2B/dataforb2b-mcp", "dataforb2b-mcp"],
      "env": {
        "DATAFORB2B_API_KEY": "your-api-key"
      }
    }
  }
}

Any other MCP client

Same command/args/env shape as above — every MCP client that supports stdio servers reads it the same way. Alternatively, run the server yourself and point your client at the process:

DATAFORB2B_API_KEY=your-api-key uvx --from git+https://github.com/DataForB2B/dataforb2b-mcp dataforb2b-mcp

or with the flag instead of the env var:

uvx --from git+https://github.com/DataForB2B/dataforb2b-mcp dataforb2b-mcp --api-key your-api-key

Development

git clone https://github.com/dataforb2b/dataforb2b-mcp.git
cd dataforb2b-mcp
uv sync
DATAFORB2B_API_KEY=your-api-key uv run dataforb2b-mcp

Inspect the server interactively with the MCP Inspector:

DATAFORB2B_API_KEY=your-api-key npx @modelcontextprotocol/inspector uv run dataforb2b-mcp

License

MIT

Available Tools

6 tools
enrich_companyCompany EnrichmentA
Read-onlyIdempotent

Enrich a single company with full data from public sources (description, industry, headquarters, employee count, funding, offices, etc.).

Inputs: company_identifier (required). Returns: error, detail, company

enrich_profileProfile EnrichmentA
Read-onlyIdempotent

Enrich a single professional profile with detailed profile data, work email, personal email, and/or GitHub profile. Each enrichment flag controls one data source and bills separately.

Inputs: profile_identifier (required), enrich_github, enrich_profile, enrich_work_email, +1 more. Returns: error, phone, detail, profile, +3 more

search_companySearch CompaniesA
Read-onlyIdempotent

Search companies using structured filters.

Use this for any company search: translate the user's intent (e.g. "AI startups in France that raised Series A") into structured filters on specific columns (industry, country, employee count range, funding stage, founded year, etc.). The available columns and operators are documented on the filters parameter.

Returns paginated companies matching ALL/ANY filter conditions depending on the op of the FilterGroup.

Inputs: filters (required), count, offset, order_by, +2 more. Returns: count, error, total, detail, +4 more

search_jobsSearch Job PostingsA
Read-onlyIdempotent

Search job postings on LinkedIn or Indeed by keyword, location, and filters (employment type, remote/onsite, seniority, freshness, employer). Useful for hiring signals: which companies are recruiting for what roles, where.

Inputs: count, offset, country, keyword, +10 more. Returns: count, error, total, detail, +4 more

search_peopleSearch PeopleA
Read-onlyIdempotent

Search professional profiles using structured filters.

Use this for any people search: translate the user's intent (e.g. "AI engineers in Berlin who recently raised") into structured filters on specific columns (country, company size range, funding stage, etc.). The available columns and operators are documented on the filters parameter.

Returns paginated profiles matching ALL/ANY filter conditions depending on the op of the FilterGroup.

To find people at a SPECIFIC company you already identified (e.g. via search_company or enrich_company), filter on current_company_id (or past_company_id) with that company's id (e.g. "org_xxx") — NOT current_company (the name), which also matches other companies that happen to share the name and adds noise.

To target people by their COMPANY's business (e.g. "founders of AI sales-agent startups"), qualify the company itself: filter current_company_category (lowercase, holds precise niche values — "artificial intelligence", "sales automation", "saas") and/or current_company_industry, and when no category fits the niche, current_company_keyword (full-text on the employer's name/tagline/description — OR several "=" phrase variants). keyword searches the PERSON's own headline, so it fits attributes of the person themselves; for a company trait it matches unrelated people ("sales agent" → real-estate agents, sales reps).

Inputs: filters (required), count, offset, order_by, +3 more. Returns: count, error, total, detail, +5 more

search_postsSearch Social PostsA
Read-onlyIdempotent

Search social posts by keyword on LinkedIn, Twitter/X, or Reddit — or target a single post by URL. Great for intent signals: find people talking about a topic, then use include to attach the post's engagers (each reactor / commenter is a lead, with their profile).

Two modes:

  1. Keyword search: set keyword (+ optional filters).

  2. Single post: set post_url (with include to get its engagers).

Inputs: count, match, offset, include, +13 more. Returns: count, error, total, detail, +4 more

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.4/5.0

    Scored across 6 tools

    Disambiguation5/5

    Each tool targets a distinct resource type (people, companies, posts, jobs) or a distinct action (search vs. enrich). The descriptions clearly differentiate search_people from enrich_profile and search_company from enrich_company, with no overlapping purposes.

    Naming Consistency5/5

    All tool names follow a consistent verb_noun pattern: search_* for discovery and enrich_* for data augmentation. This predictable structure makes it easy to infer function from the name alone.

    Tool Count5/5

    With 6 tools, the set is well-scoped for a B2B data platform. Each tool covers a distinct data operation (search people, companies, posts, jobs; enrich profiles, companies) without redundancy or bloat.

    Completeness5/5

    The tool surface covers the full expected lifecycle for a B2B data provider: discovery via search across all major entity types, and enrichment for deeper data on individuals and companies. No obvious dead ends or missing critical operations like retrieving by ID or exporting results.

    Maintenance

    ActivityInactive
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      D
      maintenance
      Give your AI agent access to 60M+ companies and 300M+ verified contacts. Enrich leads, find work emails, discover tech stacks, and identify buying intent — directly from Claude, Cursor, Windsurf, or any MCP-compatible AI agent.
      11
      9 npm
      MIT
    • F
      license
      B
      quality
      D
      maintenance
      Exposes Apollo.io API functionalities as MCP tools for people and organization enrichment, search, and job postings. Enables natural language interaction with Apollo.io data.
      5
      16
      -
    • F
      license
      Not graded
      quality
      D
      maintenance
      Enables B2B lead intelligence through seven tools for company lookup, domain enrichment, person search, email finding, and startup discovery, accessible via MCP for AI agents.
      -
    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables AI agents to search B2B people and companies, enrich contacts with verified emails and mobile phone numbers via waterfall enrichment across 25+ providers, and export results as CSV or JSON. Supports prospecting, org mapping, candidate sourcing, and outreach workflows through OAuth-authenticated remote access from any MCP-compatible client.
      14
      MIT