pipedrive-agent-cli

CLI and MCP server for the Pipedrive CRM API. Manage deals, persons, organizations, activities, leads, notes, pipelines, products, webhooks, and more — from the command line or via AI agents.

Install

npm install -g pipedrive-agent-cli

Authentication

Get your API token from Pipedrive: Settings > Personal Preferences > API

# Option 1: Environment variable (recommended for agents)
export PIPEDRIVE_API_TOKEN="your_token_here"

# Option 2: Interactive login (saves to ~/.pipedrive-cli/config.json)
pipedrive login

# Option 3: Per-command flag
pipedrive deals list --api-key "your_token_here"

Resolution order: --api-key flag > PIPEDRIVE_API_TOKEN env var > stored config.

Usage

pipedrive deals list --pretty
pipedrive deals create --title "New Deal" --value 50000 --currency USD
pipedrive deals update 123 --status won
pipedrive persons search --term "Jane"
pipedrive activities create --subject "Follow-up call" --type call --deal-id 123
pipedrive users me --fields data.name,data.email

Output

All commands output JSON to stdout. Options:

| Flag | Effect | |------|--------| | --pretty | Pretty-printed JSON | | --fields id,title,value | Select specific fields | | --quiet | Suppress output, exit codes only |

Errors go to stderr as {"error":"message","code":"ERROR_CODE"} with exit code 1.

Commands (44 tools)

| Group | Commands | |-------|----------| | deals | list, get, create, update, delete, search, activities, persons | | persons | list, get, create, update, delete, search, deals | | organizations | list, get, create, update, delete, search, deals, persons | | activities | list, get, create, update, delete | | leads | list, get, create, update, delete, search | | notes | list, get, create, update, delete | | pipelines | list, get, deals | | stages | list, get | | users | me, list, get | | search | items, field | | webhooks | list, create, delete | | products | list, get, create, update, delete, search |

Run pipedrive <group> <command> --help for detailed usage per command.

MCP Server

For AI agent integrations via the Model Context Protocol:

# Start MCP server (stdio transport)
pipedrive mcp

Claude Code config (~/.claude/settings.json):

{
  "mcpServers": {
    "pipedrive": {
      "command": "pipedrive",
      "args": ["mcp"],
      "env": { "PIPEDRIVE_API_TOKEN": "your_token_here" }
    }
  }
}

All 44 commands register as MCP tools with names like deals_list, persons_create, webhooks_delete.

API Details

• API version: Pipedrive REST API v1 (https://api.pipedrive.com/v1)
• Auth method: x-api-token header
• Response format: All responses wrapped in { success: true, data: ... }
• Pagination: Offset-based (--start, --limit)
• Rate limits: Token-based budget (~30K tokens/day per seat). Auto-retry on 429 with exponential backoff.
• Update methods: PUT for most resources; PATCH for leads
• Leads: Use UUID string IDs (not integers). Value is a nested { amount, currency } object — use --value and --currency flags together.
• Webhook event actions: create | change | delete | *
• Webhook event objects: deal | person | organization | activity | lead | product | note | pipeline | stage | user | *

Agent Guide

See AGENTS.md for comprehensive AI agent usage — workflow examples, skill patterns, and tips for automation.

License

MIT