mcpField
M
mcp-apollo

mcp-apollo

Local
github.com/AgenTeam-AI-2026Business & Marketing0Updated 29w ago

About mcp-apollo

mcp-apollo is an open-source, self-hosted MCP server for sales teams and agents using Apollo.io data. It provides people and company search and enrichment while running on the deployer's Cloudflare account and passing the API key per request.

Key features

  1. 01Search people by role, location, and company
  2. 02Enrich individual profiles
  3. 03Search and enrich company profiles
  4. 04Retrieve company job postings
  5. 05Find verified email by Apollo person ID
  6. 06Return rate-limit metadata

Use cases

  1. 01Build prospect lists from company and role filters
  2. 02Research a person or company before outreach
  3. 03Identify hiring activity at a target company
  4. 04Retrieve a person's verified email

MCP support

The server exposes Apollo.io search, enrichment, job-posting, and email lookup operations as MCP tools. It runs as a Cloudflare Worker deployed by the user and receives the Apollo API key with each request without storing it.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

mcp-apollo

An open-source, self-hosted MCP server for Apollo.io — gives your AI agents direct access to Apollo's B2B sales intelligence database.

Deploy it to your own Cloudflare account in under 5 minutes. You own the infrastructure, you control the costs. No shared hosting, no vendor lock-in.

Built and maintained by AgenTeam. MIT licensed.


How it works

Your Cloudflare Workers account
  └── mcp-apollo Worker  (your deploy, your infra)
        └── MCP client sends requests with your Apollo API key
              └── Worker proxies to Apollo.io API
                    └── Returns enriched data to your agent
  • You deploy the Worker to your own Cloudflare account (free tier handles most use cases)
  • You pass your own Apollo.io API key on every request — the Worker never stores it
  • No shared infrastructure — every user runs their own isolated instance

Quick deploy (5 minutes)

Prerequisites
Steps
# 1. Clone the repo
git clone https://github.com/AgenTeam-AI-2026/mcp-apollo
cd mcp-apollo

# 2. Install dependencies
npm install

# 3. Log in to your Cloudflare account
npx wrangler login

# 4. Deploy
npx wrangler deploy

Your server is now live at:

https://mcp-apollo.<your-subdomain>.workers.dev

That's your personal MCP server URL. Use it in the connection configs below.


Connect your MCP client

Replace YOUR_WORKER_URL with your deployed URL and YOUR_APOLLO_KEY with your Apollo.io API key.

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "apollo": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://YOUR_WORKER_URL/mcp"],
      "env": {
        "APOLLO_API_KEY": "YOUR_APOLLO_KEY"
      }
    }
  }
}

Restart Claude Desktop after saving.

Claude Code
claude mcp add apollo \
  --transport http \
  --url https://YOUR_WORKER_URL/mcp \
  --header "Authorization: Bearer YOUR_APOLLO_KEY"
Cursor

In .cursor/mcp.json:

{
  "mcpServers": {
    "apollo": {
      "url": "https://YOUR_WORKER_URL/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_APOLLO_KEY"
      }
    }
  }
}
Windsurf

In ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "apollo": {
      "serverUrl": "https://YOUR_WORKER_URL/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_APOLLO_KEY"
      }
    }
  }
}

Get an Apollo.io API key

  1. Log in to apollo.io
  2. Go to Settings → Integrations → API
  3. Click Create API Key
  4. Copy the key — paste it into your MCP client config above

Tools

apollo_search_people

Search Apollo's database for people matching filters.

ParameterTypeDescription
person_titlesstring[]Job titles, e.g. ["VP Engineering", "CTO"]
person_senioritiesstring[]"vp", "c_suite", "director", "manager", "senior"
organization_domainsstring[]Company domains, e.g. ["stripe.com"]
person_locationsstring[]Locations, e.g. ["New York", "San Francisco"]
organization_num_employees_rangesstring[]Size ranges, e.g. ["51,200", "201,500"]
pagenumberPage number (default: 1)
per_pagenumberResults per page, max 25 (default: 10)

apollo_enrich_person

Get a full enriched profile for a specific person.

ParameterTypeDescription
emailstringMost reliable identifier
linkedin_urlstringLinkedIn profile URL
first_name + last_name + organization_namestringName + company combo

At least one identifier is required.


apollo_search_companies

Search for companies matching criteria.

ParameterTypeDescription
q_organization_keyword_tagsstring[]Keywords/industries, e.g. ["SaaS", "fintech"]
organization_locationsstring[]HQ locations
organization_num_employees_rangesstring[]Size ranges
organization_funding_stagesstring[]"Seed", "Series A", "Series B" etc.
pagenumberPage (default: 1)
per_pagenumberMax 25 (default: 10)

apollo_enrich_company

Get a full company profile by domain or name.

ParameterTypeDescription
domainstringe.g. "stripe.com" — preferred
namestringCompany name — fallback

Returns: description, headcount, founding year, total funding, funding stage, tech stack, keywords.


apollo_get_job_postings

Get active job postings for a company — a strong buying/hiring signal.

ParameterTypeDescription
organization_idstringApollo org ID from apollo_enrich_company
job_titlesstring[]Optional filter, e.g. ["engineer", "sales"]

apollo_find_email

Retrieve a person's verified email by their Apollo person ID.

ParameterTypeDescription
person_idstringApollo person ID from apollo_search_people

Apollo.io rate limits

PlanRequests / hour
Free50
Basic200
Professional1,000
OrganizationCustom

The server returns rate limit metadata with every response (rate_limits.remaining, rate_limits.resetAt). On 429 errors it returns an informative message — it never silently fails.


Cloudflare Workers free tier

MetricFree limit
Requests100,000 / day
CPU time10ms / request

100k requests/day is enough for most individual or small-team deployments. If you exceed it, upgrade to the Workers Paid plan ($5/month for 10M requests).


Local development

npm install
npm run dev
# Server runs at http://localhost:8787

Run tests

npm test                                        # unit + RALPH-loop tests
APOLLO_API_KEY=your_key npx vitest run test/e2e # E2E against live Apollo API

CI/CD on your fork

The repo includes GitHub Actions workflows:

  • ci.yml — runs typecheck, build check, and tests on every push
  • deploy.yml — deploys to Cloudflare Workers on merge to main
  • e2e.yml — nightly E2E tests against the live Apollo API

To enable auto-deploy, add these secrets in your fork under GitHub → Settings → Secrets → Actions:

SecretWhere to get it
CLOUDFLARE_API_TOKENCloudflare Dashboard → My Profile → API Tokens → Create Token (use "Edit Cloudflare Workers" template)
CLOUDFLARE_ACCOUNT_IDCloudflare Dashboard → Workers & Pages → right sidebar

To enable nightly E2E tests, also add APOLLO_API_KEY as a secret and set the repo variable APOLLO_E2E_ENABLED to true.


Security

  • Your Apollo API key is passed as a Bearer token in the Authorization header on each request
  • The Worker never stores, logs, or persists your API key
  • Each request is stateless — the key is used only for that request's Apollo API call
  • You control the entire deployment — no data touches third-party infrastructure

Built by AgenTeam

AgenTeam builds AI agent teams for B2B sales and GTM operations. mcp-apollo is one of several open-source MCP servers we publish to give agents direct access to the tools modern sales teams rely on.


License

MIT © AgenTeam-AI-2026

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.