docs: update pipedream docs

sources/platform/integrations/index.mdx

@@ -237,6 +237,13 @@ Trigger Apify Actors from workflow automation platforms (n8n, Make, and Zapier),
         imageUrl="/img/platform/integrations/gumloop.svg"
         smallImage
     />
+    <Card
+        title="Pipedream"
+        desc="Run Apify Actors and react to Actor or task run events."
+        to="/integrations/pipedream"
+        imageUrl="/img/platform/integrations/pipedream.svg"
+        smallImage
+    />
 </CardGrid>
 
 [See all Workflow and notification integrations →](/integrations/workflows-and-notifications)

sources/platform/integrations/workflows-and-notifications/index.mdx

@@ -39,6 +39,13 @@ Trigger Apify Actors from workflow automation platforms (n8n, Make, and Zapier),
         imageUrl="/img/platform/integrations/gumloop.svg"
         smallImage
     />
+    <Card
+        title="Pipedream"
+        desc="Developer-first workflow automation. Run Apify Actors and react to Actor or task run events."
+        to="/integrations/pipedream"
+        imageUrl="/img/platform/integrations/pipedream.svg"
+        smallImage
+    />
 </CardGrid>
 
 For the full list, see the **Workflows and notifications** section in the sidebar.

sources/platform/integrations/workflows-and-notifications/pipedream.md

@@ -1,6 +1,6 @@
 ---
 title: Pipedream integration
-description: Connect Apify with Pipedream to automate workflows by running Actors, managing storage, and reacting to Actor or task run events.
+description: Connect Apify with Pipedream to automate workflows by running Actors, scraping single pages, managing storage, and reacting to Actor or task run events.
 sidebar_label: Pipedream
 slug: /integrations/pipedream
 ---
@@ -23,9 +23,9 @@ Before you begin, make sure you have:
 1. Log into your Pipedream account and [create a new workflow](https://pipedream.com/docs/workflows).
 1. [Add an Apify step](https://pipedream.com/docs/workflows/building-workflows/steps) (trigger or action) to your workflow. Select one of two Apify apps:
 
-    ![Selecting the Apify app in Pipedream](../images/pipedream-select-app.png)
+    ![Selecting the Apify app in Pipedream](../images/pipedream/pipedream-select-app.webp)
 
-    - **Apify** - Authenticate with your Apify API token. Find it in [Apify Console](https://console.apify.com/settings/integrations) under **Settings > Integrations**.
+    - **Apify** - Authenticate with your Apify API token. Find it in [Apify Console](https://console.apify.com/settings/integrations) under **Settings > API & Integrations**.
     - **Apify (OAuth)** - Authorize access to your Apify account via OAuth.
 1. Follow the prompts to authenticate your account.
 
@@ -38,10 +38,10 @@ Before you begin, make sure you have:
 
 1. [Create a new workflow](https://pipedream.com/docs/workflows) in Pipedream.
 1. Select **Add Trigger** and search for **Apify**.
-1. Select the trigger you want to use, e.g. **New Finished Actor Run**.
-1. Configure the trigger by selecting the Actor or task to monitor.
+1. Select the trigger you want to use, e.g. **New finished Actor run (instant)**.
+1. Configure the trigger by selecting the Actor or task to monitor. Leave **Trigger on run states** empty to fire on every terminal state (**Succeeded**, **Failed**, **Timed out**, and **Aborted**), or pick the states you care about.
 
-    ![Configuring an Apify trigger in Pipedream](../images/pipedream-trigger.png)
+    ![Configuring an Apify trigger in Pipedream](../images/pipedream/pipedream-trigger.webp)
 1. Add subsequent steps to process the output.
 
 ## Use Apify as an action
@@ -52,29 +52,102 @@ Before you begin, make sure you have:
 1. Click **+** to add a step and search for **Apify**.
 1. Select the action you want to use, e.g. **Run Actor**.
 1. Configure the action parameters:
-    - Select the Actor from Apify Store or your recently used Actors
-    - Provide the Actor input as JSON
-    - Set optional parameters such as timeout, memory, and build tag
-
-    ![Configuring an Apify action in Pipedream](../images/pipedream-action.png)
-1. Add another Apify step with **Get Dataset Items** to retrieve the Actor's output.
+    - Set **Search Actors from** to **Apify Store Actors** or **Recently used Actors**, then pick the Actor.
+    - Fill in the Actor's input fields, which are generated from the Actor's input schema. Actors without an input schema get a single **Properties** field that accepts raw JSON.
+    - Leave **Wait for Finish** set to `true` (the default) to wait for the run, or set it to `false` to return run details immediately.
+    - With `true`, the step returns the key-value store record named by **Output Record Key** (`OUTPUT` by default). It does not return dataset items.
+    - Set optional fields as needed: **Build** (a build tag or build number), **Timeout (seconds)**, **Memory (MB)** (powers of two from 128 MB to 32 GB), **Max Items**, **Max Total Charge USD**, and **Webhook URL**.
+
+    ![Configuring an Apify action in Pipedream](../images/pipedream/pipedream-action.webp)
+1. Add a **Get dataset items** step to fetch the results. Most Actors write scraped rows to a dataset, not to `OUTPUT`.
 1. Add any subsequent steps to process or store the data.
 
+:::caution Building workflows with AI
+
+Pipedream's [Apify app page](https://pipedream.com/apps/apify) can generate a workflow for you with Pipedream's AI builder. Because the Apify connector is schema-driven, the AI builder can consume your Pipedream AI tokens quickly and might not configure required inputs reliably. For predictable results, add the Apify triggers and actions to your workflow directly, as described above.
+
+:::
+
+## Handle long-running Actor runs
+
+Some Actor runs can outlast a single Pipedream step.
+
+- **Run task**: waits asynchronously (webhook + 30-second polling fallback), up to one day.
+- **Run Actor**: waits inside the step and can hit step timeout.
+
+For longer runs with **Run Actor**, split across two workflows:
+
+1. In the first workflow, add **Run Actor** and set **Wait for Finish** to `false`. The step returns the run details immediately.
+1. In a second workflow, use the **New finished Actor run (instant)** trigger for the same Actor.
+1. Add **Get dataset items** after the trigger.
+
+Alternatively, save the Actor configuration as an [Actor task](/actors/running/tasks) and use **Run task** with **Wait for Finish** set to `true`, then add **Get dataset items** to fetch the results.
+
+## Handle large Actor output
+
+When **Run Actor** waits for a run, it returns the record named by **Output Record Key** (`OUTPUT` by default). The connector caps inline output at 256 KB to stay within Pipedream's step output size, so a larger record comes back as a reference object with `recordUrl`, `keyValueStoreId`, and `recordKey`.
+
+To read large output:
+
+- Fetch `recordUrl` in a later step.
+- Use **Get key-value store record** with `keyValueStoreId` and `recordKey`.
+
+If the Actor writes to a dataset, use **Get dataset items**.
+
 ## Triggers
 
-- **New finished Actor run (instant)** - Triggers when a selected Actor run finishes.
-- **New finished task run (instant)** - Triggers when a selected task run finishes.
+Both triggers register an Apify webhook when you deploy the workflow and remove it when you disable the workflow.
+
+- **New finished Actor run (instant)** - Emits an event when a run of the selected Actor finishes.
+- **New finished task run (instant)** - Emits an event when a run of the selected task finishes.
+
+Each trigger takes a **Trigger on run states** field listing the terminal run states: **Succeeded**, **Failed**, **Timed out**, and **Aborted**. Leave it empty to fire on all four, which is the default.
 
 ## Actions
 
-- **Run Actor** - Runs a selected Actor with customizable input and configuration.
-- **Run task** - Runs a selected Actor task and optionally waits for it to finish.
-- **Run task synchronously** - Runs a selected task and returns its dataset items when it finishes.
-- **Scrape single URL** - Runs a scraper on a specified URL and returns its content as HTML. Use this for extracting content from a single page, e.g. in LLM workflows.
-- **Get dataset items** - Retrieves items from a [dataset](/storage/dataset).
-- **Get key-value store record** - Retrieves a record from a [key-value store](/storage/key-value-store).
+- **Run Actor** - Runs a selected Actor and, by default, waits for it to finish and returns the `OUTPUT` record. Input fields are generated from the Actor's input schema. Optional fields: **Build**, **Timeout (seconds)**, **Memory (MB)**, **Max Items**, **Max Total Charge USD**, **Webhook URL**, and **Output Record Key**.
+- **Run task** - Runs a selected task and, by default, waits for it to finish and returns the run, without its output. Use **Override Input (JSON)** to replace the task's saved input for a single run, and leave it empty to use the saved input. Optional fields: **Build**, **Timeout (seconds)**, and **Memory (MB)**.
+- **Scrape single URL** - Runs [Website Content Crawler](https://apify.com/apify/website-content-crawler) on a specified URL and returns its content as HTML. Use this for extracting content from a single page, e.g. in LLM workflows. **Crawler Type** selects the engine: **Firefox (Headless Browser)** renders JavaScript (the default), **Cheerio (Raw HTTP)** sends raw HTTP requests and renders none, and **Adaptive** switches per page.
+- **Get dataset items** - Retrieves items from a [dataset](/storage/dataset), specified by ID or name. **Limit** and **Offset** page through the items, and **Fields**, **Omit**, **Flatten**, and **Clean** shape each item.
+- **Get key-value store record** - Retrieves a record from a [key-value store](/storage/key-value-store). A JSON record is returned as parsed fields, and any other content type as a file reference.
 - **Set key-value store record** - Creates or updates a record in a [key-value store](/storage/key-value-store).
 
+## Use Apify with AI agents (MCP)
+
+To use Apify Actors directly with AI agents and MCP-compatible clients, use the [Apify MCP server](/integrations/mcp).
+
+Pipedream also hosts an [MCP server for Apify](https://mcp.pipedream.com/app/apify). Because Apify provides its own MCP server, use the Pipedream one only if you want Apify available alongside your other Pipedream apps in a single MCP client.
+
+## Troubleshooting
+
+### The workflow times out before the Actor finishes
+
+**Run Actor** waits for the run inside the step, so a long run can exceed Pipedream's step limit. Set **Wait for Finish** to `false` and pick the results up from a second workflow, or use **Run task**, which waits asynchronously. See [Handle long-running Actor runs](#handle-long-running-actor-runs).
+
+### The Actor dropdown is empty
+
+**Search Actors from** defaults to **Recently used Actors**, which is empty on a new Apify account. Switch it to **Apify Store Actors** to browse [Apify Store](https://apify.com/store), or enter an Actor ID or an `owner~actor-name` identifier directly.
+
+### The Actor has no builds yet
+
+An Actor cannot run until it has been built. Open the Actor in [Apify Console](https://console.apify.com/actors) and build it, or run `apify push` with the [Apify CLI](/cli), then run the step again.
+
+### Apify can't find the build you entered
+
+The **Build** dropdown lists tagged builds only, because some Actors have thousands of builds. To use an untagged build, type its build number, such as `0.1.2`, into the field instead of picking from the list.
+
+### Apify rejects the memory value
+
+**Memory (MB)** must be a power of two between 128 MB and 32 GB, and **Run Actor** narrows the list further to the range the Actor's build declares. Pick a value from the dropdown rather than typing one.
+
+### The step reports invalid JSON input
+
+Actors without an input schema take a single **Properties** field, and **Run task** takes **Override Input (JSON)**. Both must contain a valid JSON object. Use `{}` to run with the saved or default input, and copy the input from the Actor's **Input** tab in Apify Console if you are unsure of the shape.
+
+### The trigger won't deploy
+
+Both triggers need a target before they can register their Apify webhook. Select an Actor for **New finished Actor run (instant)**, or a task for **New finished task run (instant)**, then deploy again.
+
 ## Resources
 
 - [Apify integration page on Pipedream](https://pipedream.com/apps/apify)