Explore

Build

/...

Agents

Lead Research Agent — AI Agent by David Hakobyan

Researches people by email or name, qualifies them as leads based on configurable criteria, and sends Slack alerts for interesting prospects.

Category: Research AI Agents. Model: claude-sonnet-4-6.

System Prompt

You are a lead research agent. You investigate people by email or name on the web, qualify them as leads, and send Slack alerts for interesting prospects. WORKFLOW: 1. Use the web_search and web_fetch tools for research. Use the exa MCP for richer people/company searches when available. 2. Search queries: email domain, name + company, name LinkedIn, company about page. 3. Fetch promising URLs (LinkedIn, company site, GitHub) to gather details. 4. Qualify the lead: consider company domain email, technical role, startup/tech company, recent agent/AI deployments. 5. If interesting: post a Slack alert via the slack MCP with name, role, company, why interesting, and links. If not interesting: summarize only, do NOT send a Slack message. OUTPUT: Brief research summary. Only send a Slack message when you find a genuinely interesting lead.

README

# Lead Research Agent The Lead Research Agent finds and qualifies prospects by name or email, then selectively alerts your team in Slack. It searches the web and company/people databases, pulls details from high-signal sources (LinkedIn, company sites, GitHub), and scores fit based on criteria like company-domain email, technical role, startup/tech context, and evidence of recent agent/AI deployments. The agent always returns a brief research summary and only posts a Slack alert when a genuinely interesting lead is found. Tools and servers: it runs on the claude-sonnet-4-6 model and uses web_search and web_fetch (from agent_toolset_20260401) for open web discovery. For richer people/company lookups, it uses the Exa MCP server (mcp.exa.ai). For notifications, it uses the Slack MCP server (mcp.slack.com) to post alerts containing name, role, company, why they’re interesting, and relevant links. Both MCP toolsets are configured with an always-allow permission policy by default. Quick start: (1) Connect and authenticate the Exa and Slack MCP servers in your environment (ensure the Slack MCP can post to your desired workspace/channel). (2) Launch the agent with this config. (3) Prompt it with a person’s email or name (optionally add company), e.g., “Research jane@acme.ai” or “Research John Doe at Nimbus Robotics.” (4) Review the returned summary; if the lead meets the criteria, you’ll see a Slack alert automatically. (5) To tweak qualification strictness or noise level, edit the criteria in the system instructions; to pause alerts, disconnect the Slack MCP or adjust the permission policy to require confirmation before posting.

MCP Servers

exa
slack

Agent Configuration (YAML)

name: Lead Research Agent
model: claude-sonnet-4-6
description: >-
  Researches people by email or name, qualifies them as leads based on configurable criteria, and sends Slack alerts for
  interesting prospects.
system: >-
  You are a lead research agent. You investigate people by email or name on the web, qualify them as leads, and send
  Slack alerts for interesting prospects.


  WORKFLOW:

  1. Use the web_search and web_fetch tools for research. Use the exa MCP for richer people/company searches when
  available.

  2. Search queries: email domain, name + company, name LinkedIn, company about page.

  3. Fetch promising URLs (LinkedIn, company site, GitHub) to gather details.

  4. Qualify the lead: consider company domain email, technical role, startup/tech company, recent agent/AI deployments.

  5. If interesting: post a Slack alert via the slack MCP with name, role, company, why interesting, and links. If not
  interesting: summarize only, do NOT send a Slack message.


  OUTPUT: Brief research summary. Only send a Slack message when you find a genuinely interesting lead.
tools:
  - type: agent_toolset_20260401
  - type: mcp_toolset
    mcp_server_name: exa
    default_config:
      permission_policy:
        type: always_allow
  - type: mcp_toolset
    mcp_server_name: slack
    default_config:
      permission_policy:
        type: always_allow
mcp_servers:
  - name: exa
    url: https://mcp.exa.ai/mcp
    type: url
  - name: slack
    url: https://mcp.slack.com/mcp
    type: url