# prospectapis-mcp

prospectapis-mcp is the MCP server for ProspectAPIs (https://prospectapis.com), for sales and prospecting teams and the AI agents they run in Claude Code, Claude Desktop, Cursor and other MCP clients. It gives your agent twenty-five tools over the ProspectAPIs API: search funding rounds by date, round, amount, company, investor and country, get new matching rounds pushed to your webhook, verify, find and bulk-check email addresses, report what bounced, and order a cited research brief on any company or person. Funding records carry the public announcement URL when there is one. Every account starts with a free $1 balance and no card: records cost $0.02, an email check $0.004 and a completed brief $0.15, with no seats and no subscription. It needs Node.js 22 or later.

Install (Claude Code): `claude mcp add prospectapis -e PROSPECTAPIS_API_KEY=pa_live_... -- npx -y prospectapis-mcp`

Other clients: run `npx -y prospectapis-mcp` with `PROSPECTAPIS_API_KEY` from https://prospectapis.com

Website: https://prospectapis.com/mcp

| Tool | Endpoint | Cost |
|---|---|---|
| `funding_search` | `GET /v1/funding` | $0.02 per record returned; no match is free |
| `funding_get` | `GET /v1/funding/:id` | $0.02 when found; a missing id is free |
| `research_submit` | `POST /v1/research` | one research at `research_price_usd` (see `account_balance`), charged only when the brief completes; refunded on failure |
| `research_get` | `GET /v1/research/:id` | free; poll until `status` is `completed` or `failed` |
| `email_verify` | `POST /v1/email/verify` | $0.004 per `valid` or `invalid` result; `catch-all` and `unknown` are free; 1 to 25 addresses a call |
| `email_verify_stats` | `GET /v1/email/verify/stats` | free |
| `email_outcomes_report` | `POST /v1/email/outcomes` | free; 1 to 100 outcomes a call |
| `email_verify_job_create` | `POST /v1/email/verify/jobs` | free to start; $0.004 per `valid` or `invalid` result as it is checked; up to 50,000 addresses |
| `email_verify_job_get` | `GET /v1/email/verify/jobs/:id` | free |
| `email_verify_job_results` | `GET /v1/email/verify/jobs/:id/results` | free; 1,000 a page |
| `watchlist_create`, `watchlist_list`, `watchlist_get`, `watchlist_deliveries`, `watchlist_test`, `watchlist_delete` | `/v1/watchlists` | free to manage; a delivered event costs one record, only when your webhook answers 2xx |
| `account_balance` | `GET /v1/account` | free |
| `feedback_send` | `POST /v1/feedback` | free; reports a bug, wrong field or missing capability to the team (10 an hour per key) |
| `pricing_get` | `GET /pricing` | free, no key needed; the live price list |
| `api_version` | `GET /v1/version` | free, no key needed |

`funding_search` filters: `since`, `until` (YYYY-MM-DD), `round` (comma list such as `seed,series_a`),
`min_amount_usd`, `max_amount_usd`, `company`, `domain`, `investor`, `country`, `limit` (1 to 100) and
`cursor` (the previous page's `next_cursor`, passed back unchanged).

`research_submit` takes exactly one prospect identifier (`domain`, `website_url`, `company_name`,
`company_linkedin_url`, a work `email` or `linkedin_url`), or `person_name` with `company_name`, plus an optional
`context` (your own product, at most 1,000 characters) and `purpose` (`account_research`, `pre_call` or
`outreach`). It returns a `research_id` at once; a brief takes 20 to 90 seconds.

`email_verify` takes `emails`, a list of 1 to 25 addresses. Each result has `status` (`valid`, `invalid`,
`catch-all` or `unknown`), `reason`, `reason_text`, `score` (0 to 100), `recommendation` (`send`, `send_with_caution`, `retry_later` or `do_not_send`), `charged`, `flags`, `smtp_code` and `catch_all_tier`. If the balance covers only
some addresses, the rest come back `unknown` with reason `not_checked_insufficient_credits` and are not charged.

## Configure

| Variable | Required | Default |
|---|---|---|
| `PROSPECTAPIS_API_KEY` | yes (all tools except `pricing_get` and `api_version`) | none |
| `PROSPECTAPIS_BASE_URL` | no | `https://api.prospectapis.com` |

Client config (for example `claude_desktop_config.json`):

```json
{
  "mcpServers": {
    "prospectapis": {
      "command": "npx",
      "args": ["-y", "prospectapis-mcp"],
      "env": { "PROSPECTAPIS_API_KEY": "pa_live_..." }
    }
  }
}
```

Claude Code: `claude mcp add prospectapis -e PROSPECTAPIS_API_KEY=pa_live_... -- npx -y prospectapis-mcp`.

## FAQ

**Where does the funding data come from?** Records carry the public announcement URL when there is one.

**What is free?** A $1 balance when you sign up (about 50 funding records, 250 email checks or 6 research briefs), any search that matches nothing, catch-all and unknown email results, watchlist management, and the balance, pricing and version tools.

**Which clients work?** Claude Code, Claude Desktop, Cursor and any client that reads an mcpServers config, through `npx -y prospectapis-mcp`.

## Develop

```bash
cd mcp
npm install
npm test        # mocked fetch over an in-memory MCP transport, no network
npm run check
```

The package is versioned independently of the API (SemVer, `CHANGELOG.md` in this folder). It is marked
`private` so it cannot be published by accident; publishing is done only from the product's own npm identity.