Pipedrive CRM

This connector uses the Pipedrive REST API.

Versions: most resources (deals, persons, organizations, activities, pipelines, stages, users) are on the newer v2 API (/api/v2/...). The universal item-search endpoint is still on v1 (/v1/itemSearch). The adapter mixes both as appropriate — you don't need to know the distinction, just pick the tool.

Setup:

1. Sign in to Pipedrive → top-right avatar → Personal preferences → API. Copy the personal API token shown there.
2. Paste it as PIPEDRIVE_API_TOKEN.
3. (Optional) If your account is on a custom company subdomain that requires it, set PIPEDRIVE_COMPANY_DOMAIN to your subdomain (e.g. acme if your CRM lives at acme.pipedrive.com). The default base URL https://api.pipedrive.com works for almost everyone since the token already routes to the right company.

Authentication: API token in query string (?api_token=...) — every request automatically appends it. The token belongs to a single user; everything the connector does is attributed to that user. For multi-user installs, use a dedicated service-account user with the permissions you want the agent to have.

Pagination: v2 endpoints use cursor-based pagination (limit + cursor). Each list response carries additional_data.next_cursor — pass it as cursor on the next call. limit defaults to 100, max 500. The adapter's list tools expose both parameters explicitly.

Search: for fuzzy text matching across multiple entity types, use pipedrive_search (universal). For type-narrowed search inside a known entity, use the per-resource pipedrive_search_deals / pipedrive_search_persons.

Custom fields: every CRM has custom fields. v2 endpoints accept and return them under the custom_fields body/response key. The shape is {field_api_key: value} — get the API keys from GET /v1/dealFields (and personFields, organizationFields). The adapter ships pipedrive_list_deal_fields / pipedrive_list_person_fields / pipedrive_list_organization_fields for this.

Rate limits: Pipedrive applies a per-token rate limit (~80 req / 2 seconds per company, plus daily quotas that scale with plan). On 429, back off — the engine does NOT auto-retry.

Webhooks not in scope: Pipedrive can push webhook events for deal/person changes; receiving them requires a separately-hosted endpoint and is outside this connector's read/write scope.

Filters & sorting: list endpoints accept filter_id, owner_id, status for deals; sort parameter accepts comma-separated field directives like add_time DESC,update_time DESC on v2. Check Pipedrive's API reference for the supported sort fields per resource.

Out of scope here: leads (v1 only), products, files, mailbox, projects, and goals — add purpose-built tools if your workflows need them.