> ## Documentation Index
> Fetch the complete documentation index at: https://docs.valendata.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Integrations: Connect Valendata to Your Existing Stack

> Connect Valendata to the tools your team already uses — push run results to any destination or trigger Skills from any platform via API.

Valendata is designed to fit inside the workflows your team already has, not replace them. Whether you want to push structured data into a database, fire a Slack alert when a run fails, or trigger a research Skill the moment a new lead lands in your CRM, there is a direct path to make it happen. Every integration falls into one of two patterns: you either push data **out** of Valendata to a destination, or you trigger Skills **from** an external tool via the API.

## Two integration patterns

Valendata supports two fundamental ways to connect with external systems:

<CardGroup cols={2}>
  <Card title="Push: send data out" icon="arrow-right-from-bracket">
    Use Workflow sinks, webhook steps, or outbound HTTP actions to deliver run output to any destination the moment a run completes. Your downstream tools receive fresh, structured JSON — no polling required.
  </Card>

  <Card title="Trigger: call Skills in" icon="arrow-right-to-bracket">
    Every published Skill is a versioned REST endpoint. Any tool that can make an HTTP request — Zapier, Make, your own backend, a cron job — can trigger a Skill and consume its output.
  </Card>
</CardGroup>

## Popular integrations

<CardGroup cols={2}>
  <Card title="Zapier" icon="bolt" href="/integrations/zapier">
    Trigger Skills from any of Zapier's 5,000+ connected apps, or use Skill output to kick off downstream Zaps. No code required.
  </Card>

  <Card title="Webhooks" icon="webhook" href="/integrations/webhooks">
    POST run result payloads to any HTTP endpoint the instant a run completes. Retry logic and HMAC-SHA256 signature validation are built in.
  </Card>

  <Card title="Postgres" icon="database">
    Sink Workflow output directly into a Postgres table. Define the table schema once and Valendata maps each run's JSON output to the right columns automatically.
  </Card>

  <Card title="Slack" icon="slack">
    Receive a Slack message whenever a run completes or fails. Wire up channel alerts in the Workflow editor so your team knows the moment fresh data is ready — or something went wrong.
  </Card>

  <Card title="HubSpot & Salesforce" icon="building">
    Enrich CRM contact and company records with live data from the web. Run a research Skill on a new deal and push the results back into HubSpot or Salesforce fields automatically.
  </Card>

  <Card title="Google Sheets" icon="table">
    Pipe scraped or extracted data into a spreadsheet. Each Skill run can append rows to an existing sheet, keeping your team's live dashboards and reports always up to date.
  </Card>

  <Card title="Notion" icon="n">
    Populate Notion databases from web sources. Schedule a Skill to run nightly and watch your Notion workspace fill with fresh, structured records — no copy-paste required.
  </Card>

  <Card title="Airtable" icon="table-cells">
    Sync live web data into your Airtable base. Map Skill output fields to Airtable columns and let Valendata keep your base current on any schedule you choose.
  </Card>
</CardGroup>

## Using the REST API as a universal connector

Every integration ultimately relies on Valendata's REST API. If your tool of choice is not listed above, you can connect it directly — any system capable of making an HTTP POST request can trigger a Skill and receive structured JSON output.

<Note>
  For tools not listed here, use the [REST API](/api-reference/skills/run) directly. Pass your API key in the `Authorization: Bearer` header, POST your parameters to `https://api.valendata.com/v1/skills/run`, and handle the JSON response however your system needs. If a tool supports custom HTTP requests — including low-code platforms, internal dashboards, or backend services — it already works with Valendata.
</Note>