Pipedrive Leads MCP. Manage your entire sales pipeline without leaving your chat.

This server hooks your Pipedrive CRM straight into your AI client. You won't have to jump between browser tabs or manually update anything in the sales funnel again. Your agent gets full visibility across every single lead, letting you manage records—creating them, updating them, and cleaning up junk—just by giving simple instructions.

Building Leads and Records: When you need a new prospect in Pipedrive, your agent uses pd_create_lead to build the record. It handles all the initial data points for that brand-new lead right away. If a lead changes status or needs a name fix, you tell it what's wrong, and the agent executes pd_update_lead, changing specific details like the stage or updating the person's name on an existing account.

Cleaning Up: Got leads that are dead ends? You don't wanna keep 'em cluttering up your pipeline. Your agent runs pd_delete_lead to remove old records from Pipedrive entirely, keeping your data clean and actionable.

Checking the Status Quo: Wanna see what's going on with all your prospects right now? You ask for a list of current leads, and the agent fires off pd_list_leads. This gives you an immediate overview and status check on multiple leads at once. If you need to dig deep into one specific prospect—say, you gotta review every detail about Lead ID 54321—the agent uses pd_get_lead to pull all the granular information for that single lead.

Managing Metadata: Before you update anything, you gotta make sure your tags and sources are right. You can check out what categories of labels exist by calling pd_lead_labels, which lists every available label category you've set up in Pipedrive. Similarly, if you wanna know where incoming leads are coming from, the agent uses pd_lead_sources to retrieve a full list of defined sources.

This lets your AI client verify proper tagging before it makes any changes to records.

Workflow Examples: You can tell your agent: 'Pull all current leads, then for the ones labeled 'Hot' that are coming from 'Website', update their stage to 'Qualified'.'

The server manages this whole sequence. It first runs pd_list_leads to gather the necessary status data; it checks available categories with pd_lead_labels; and once everything lines up, it executes pd_update_lead for every record that matches your criteria.

You can't just update names or stages; you gotta get the metadata straight first. If you don't know what labels are valid, calling pd_lead_labels stops you from using a typo and messing up your data structure. It makes sure everything stays tight.

For example, if a lead is coming in through an old channel, you can run pd_lead_sources to see the exact source name needed before running pd_update_lead. If that specific ID isn't found or needs more info, calling pd_get_lead pulls every single detail point so your agent knows exactly what it’s dealing with.

This eliminates guesswork.

It keeps everything organized. You tell the agent to build a lead, and it uses pd_create_lead. If that lead turns out to be irrelevant after review, you just say delete, and it runs pd_delete_lead. That's how easy this is. It handles the whole cycle: creation via pd_create_lead, status check via pd_list_leads or deep dive with pd_get_lead, organization using pd_lead_sources and pd_lead_labels, modification with pd_update_lead, and cleanup with pd_delete_lead.

You'll stop wasting time clicking around Pipedrive. Your AI client does the heavy lifting for you.