Encuentra y verifica contactos B2B en tiempo real. Extrae información de sitios web de empresas, Discord y Telegram, y luego valida cada correo electrónico mediante SMTP. Construido como un servidor MCP nativo, para que Claude, ChatGPT, Cursor y cualquier agente de IA puedan usarlo directamente.
Instálalo con un solo comando:
uvx --from git+https://github.com/axelfreeman/tapac-mcp tapac-mcpLuego pregúntale a tu agente: "Encuentra 20 VP de Ventas en empresas SaaS de EE. UU. con 50–500 empleados."
Las bases de datos se vuelven obsoletas. El 23% de los contactos cambia de trabajo cada año (ZoomInfo 2025), el 40% de los correos electrónicos muere en 2 años (NeverBounce). Compra una lista de contactos y una cuarta parte estará desactualizada antes de enviar un solo correo.
TAPAC no vende una base de datos. Extrae información en vivo y valida en el momento de la solicitud:
TAPAC | Bases de datos estáticas (Apollo, ZoomInfo, Hunter) | |
Fuente de datos | Extracción en tiempo real | Instantánea congelada |
Tasa de rebote | 2–5% | 10–35% |
Validación | SMTP, en el momento | Ninguna / posterior |
MCP / Agentes de IA | ✅ nativo | ❌ |
Discord / Telegram | ✅ | ❌ |
Precio | Pago por uso, $0.10–0.50/contacto | $34–$15,000+/mes o /año |
Related MCP server: Bytemine MCP Server - Contact Search & Enrich
uvx --from git+https://github.com/axelfreeman/tapac-mcp tapac-mcpRequiere uv (instalador de una línea: curl -LsSf https://astral.sh/uv/install.sh | sh).
El mismo command + args funciona en todas partes. Elige tu agente:
Claude Desktop — claude_desktop_config.json:
{
"mcpServers": {
"tapac": {
"command": "uvx",
"args": ["--from", "git+https://github.com/axelfreeman/tapac-mcp", "tapac-mcp"]
}
}
}Claude Code (terminal):
claude mcp add tapac -- uvx --from git+https://github.com/axelfreeman/tapac-mcp tapac-mcpCodex (OpenAI) — ~/.codex/config.toml:
[mcp_servers.tapac]
command = "uvx"
args = ["--from", "git+https://github.com/axelfreeman/tapac-mcp", "tapac-mcp"]Cursor — .cursor/mcp.json:
{
"mcpServers": {
"tapac": {
"command": "uvx",
"args": ["--from", "git+https://github.com/axelfreeman/tapac-mcp", "tapac-mcp"]
}
}
}Windsurf — ~/.codeium/windsurf/mcp_config.json (mismo JSON que Cursor).
DeepSeek — la aplicación de chat no registra servidores MCP locales de forma nativa. Usa DeepSeek a través de un cliente compatible con MCP (Claude Code / Cursor con la API de DeepSeek), o conéctate al endpoint alojado https://tapacapi.com/mcp/sse (próximamente).
La primera ejecución te redirige a https://tapacapi.com/get-key. Configúrala una vez:
export TAPAC_API_KEY=your_key_here"Encuentra 20 VP de Ventas en empresas SaaS de EE. UU. con 50–500 empleados, verifica los correos."
El agente pregunta los criterios que necesita, llama a tapac_find_contacts y devuelve contactos verificados (nombre, cargo, empresa, correo electrónico, fuente, estado de verificación).
100 búsquedas gratuitas, sin necesidad de tarjeta de crédito.
tapac_find_contactsEncuentra y verifica contactos B2B.
Parámetro | Tipo | Descripción |
| string | Vertical, ej. "SaaS", "healthcare", "fintech" |
| string[] | Roles, ej. |
| string | "50-500 employees", "startup", "enterprise" |
| string | "US", "Europe", "DACH", "remote" |
| string |
|
| int | Cuántos contactos (por defecto 10) |
tapac_statusVersión del servidor + estado de la clave API.
Beta. La instalación, la interfaz de herramientas y la incorporación de la clave API están activas. El motor completo de extracción → validación SMTP está en proceso de integración; se lanzará próximamente. Hasta entonces, tapac_find_contacts captura tu objetivo y devuelve una respuesta beta clara, para que puedas instalar y explorar el producto hoy mismo.
Sitio web: https://tapacapi.com
Guía de investigación profunda: https://tapacapi.com/deep-research-guide.html
Kit de herramientas gratuito (autoalojado): https://github.com/axelfreeman/b2b-contact-mining-kit
Autor: Axel Freeman
MIT.
Find and verify B2B business contacts in real time.
Ask the user for these criteria before calling:
industry: vertical, e.g. "SaaS", "healthcare", "fintech"
job_titles: roles, e.g. ["VP Sales", "CTO", "Head of Growth"]
company_size: e.g. "50-500 employees", "startup", "enterprise"
location: geography, e.g. "US", "Europe", "DACH", "remote"
source: where to search — "website", "telegram", or "discord"
limit: how many contacts to return (default 10)
Returns verified contacts (name, title, company, email, source, verification status) as structured data.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | ||
| source | No | website | |
| industry | No | ||
| location | No | ||
| job_titles | No | ||
| company_size | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
While no annotations are provided, the description adds significant context by specifying that contacts are returned as structured data with fields like name, title, company, email, source, and verification status. It also discloses that contacts are verified, which is a behavioral trait. However, it does not mention authentication needs, rate limits, or what happens when no contacts are found—minor gaps that prevent a top score.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is succinct and well-structured, with a concise opening sentence followed by a bulleted list of criteria to ask for. Every sentence provides essential information without repetition or fluff. The structure is front-loaded with the main purpose, then details, making it easy for an AI agent to parse quickly.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's moderate complexity with 6 parameters, no required fields, and no annotations, the description covers the critical aspects: purpose, required user input, parameter semantics, and return structure. The presence of an output schema partially reduces the need to describe return values further. However, a note on error handling or empty results would enhance completeness for the agent, preventing it from being a 5.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The description adds substantial meaning beyond the input schema by defining each parameter with concrete examples (e.g., industry 'SaaS', location 'US'), which compensates for the 0% schema description coverage. For the limit parameter with default 10, it clarifies the purpose. This goes well beyond the bare schema definitions, making the parameters actionable for the agent.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: 'Find and verify B2B business contacts in real time.' The verb 'find and verify' combined with the resource 'B2B business contacts' and modifier 'in real time' makes the action explicit. It also distinguishes itself from the sibling tool 'tapac_status' by focusing on contact discovery rather than status checking.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides explicit usage guidelines by instructing the user to ask for specific criteria before calling the tool, listing each parameter with examples (e.g., industry 'SaaS', job_titles '[VP Sales, CTO, ...]'). This guidance clearly tells the agent when to use the tool—when it needs to collect these details from the user—and implicitly distinguishes it from the sibling tool 'tapac_status' which likely handles status queries.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Check TAPAC server status, version, and API-key state.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description indicates the tool performs a read-only check of status, version, and API-key state. While this implies safe, non-destructive behavior, there are no annotations to confirm. The description does not elaborate on potential latency, network dependencies, or error states, which would improve transparency.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is extremely concise, using a single sentence that effectively communicates the three aspects the tool checks: server status, version, and API-key state. There is no extraneous text, making it easy for an AI agent to quickly interpret its purpose.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool has no parameters and the sibling tool is clearly for contact lookup, the description is largely sufficient for a status-check tool. However, the description does not explain the format or structure of the output, which could be important for an agent to interpret results, even though an output schema exists separately.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema has zero parameters and has 100% coverage. Since no parameters exist, there is no need for the description to add parameter-level details, and a baseline score of 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool checks TAPAC server status, version, and API-key state. It uses specific nouns and verbs to describe the function, differentiating it from sibling tools that likely interact with contacts rather than checking server status.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage for health-check or diagnostic purposes, but does not explicitly state when to use this tool versus the sibling tapac_find_contacts. It lacks guidance on prerequisites or conditions under which a status check is appropriate.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.0tapac_find_contactstapac_statusScored across 2 tools
The two tools serve completely distinct purposes: one checks server status, the other finds contacts. There is no overlap or ambiguity.
Both tools follow a consistent snake_case pattern with the prefix 'tapac_', one describing a state (status) and the other an action (find_contacts). The naming is uniform.
With only two tools, the server feels thin for a contact-finding service, but each tool is clearly defined and serves a necessary function. It is borderline but not excessive.
The server covers the core operation of finding and verifying contacts with extensive filtering, plus a status check. Minor gaps like retrieving a specific contact by ID are absent, but the main workflow is complete.
Managed LinkedIn MCP server for AI agents: search, connect, message and enrich on accounts you own.
LinkedIn outreach MCP server — 19 tools for AI agents to prospect, sequence, and manage contacts.
Agent-native MCP server over 49M+ US public and government records, privacy-first, always current.
API-first CRM for LLMs - contacts, companies, deals and activities over a native MCP server.
