AI导哈

领英外联自动化MCP服务器

Prosp MCP Server 通过模型上下文协议以编程方式管理领英外联流程，提供线索管理、营销活动、消息发送与分析功能，适用于销售自动化和社交外联场景。

  • by AvolveAi
  • Python
  • 本地部署
  • 1 星标
  • 更新于
在 GitHub 查看 AvolveAi/prosp-mcp

使用说明

Prosp MCP Server

MCP server for Prosp.ai LinkedIn outreach automation. Manage leads, campaigns, messaging, and analytics programmatically via the Model Context Protocol.

Tools (17)

Leads (8 tools)

Tool Description Type
add_lead Add a lead to a contact list and campaign write
add_lead_to_list Add a lead to a contact list only write
add_existing_lead_to_campaign Add an existing lead to a campaign write
remove_lead_from_campaign Remove a lead from a campaign destructive
delete_lead Permanently delete a lead from workspace destructive
get_leads_in_campaign Get all leads in a campaign read-only
get_lead_stage Get current stage of a lead in a campaign read-only
check_api_key Verify API key and connectivity read-only

Campaigns (5 tools)

Tool Description Type
get_all_campaigns List all campaigns in workspace read-only
get_campaign_status Get status of a campaign read-only
start_campaign Start a campaign write
stop_campaign Stop a running campaign destructive
get_analytics Get analytics for campaign or workspace read-only

Messaging (3 tools)

Tool Description Type
send_message Send a LinkedIn message to a profile write
send_voice_message Send a LinkedIn voice message write
get_conversation Get conversation history with a lead read-only

System (1 tool)

Tool Description Type
get_server_info Server version, config, loaded tools read-only

Setup

Environment Variable

export PROSP_API_KEY="your-api-key-here"

Get your API key from Prosp Settings.

Claude Desktop / Claude Code

Add to your MCP config:

{
  "mcpServers": {
    "prosp": {
      "command": "prosp-mcp",
      "env": {
        "PROSP_API_KEY": "your-api-key-here"
      }
    }
  }
}

Lazy Loading

Load only specific tool categories via TOOL_CATEGORIES:

{
  "mcpServers": {
    "prosp": {
      "command": "prosp-mcp",
      "env": {
        "PROSP_API_KEY": "your-api-key-here",
        "TOOL_CATEGORIES": "leads,campaigns"
      }
    }
  }
}

Available categories: leads, campaigns, messaging

Install

pip install prosp-mcp

Or from source:

git clone https://github.com/AvolveAi/prosp-mcp.git
cd prosp-mcp
pip install -e .

Development

pip install -e ".[dev]"
pytest tests/

API Coverage

Wraps the Prosp API v1 endpoints:

Endpoint Tool
POST /api/v1/leads add_lead
POST /api/v1/leads/list add_lead_to_list
POST /api/v1/leads/campaign add_existing_lead_to_campaign
POST /api/v1/leads/campaign/remove remove_lead_from_campaign
POST /api/v1/leads/delete delete_lead
POST /api/v1/leads/campaign/list get_leads_in_campaign
POST /api/v1/leads/stage get_lead_stage
POST /api/v1/campaigns get_all_campaigns
POST /api/v1/campaigns/status get_campaign_status
POST /api/v1/campaigns/start start_campaign
POST /api/v1/campaigns/stop stop_campaign
POST /api/v1/analytics get_analytics
POST /api/v1/messages/send send_message
POST /api/v1/messages/voice send_voice_message
POST /api/v1/conversations get_conversation

Note: Endpoint paths are inferred from Prosp's API documentation and Zapier/Make.com integration. Some paths may need adjustment once the full OpenAPI spec is available.

License

MIT

相关 MCP