Skip to Content
AutomationsActionsData

Search & Data Actions

Searching the Knowledge Base

What it does: Searches your company Knowledge Base and returns relevant chunks or documents. Use when the automation needs to answer questions or use internal content.

Inputs — Query (required), optional filters (e.g. data source, limit). Outputs — List of results (e.g. title, content, source, score) for use in prompts or summaries. See your step configuration for the exact output schema.

What it does: Searches the web for a query.

Inputs

  • Query (required) — The search string (e.g. “best practices for X”, “site:example.com”). Can include variables.
  • Max results (optional) — Maximum number of results to return (e.g. 10, 20). Caps API usage and response size.
  • Region (optional) — Region or country for results (e.g. us, uk). Affects ranking and local results.

Outputs

A list of search results (title, URL, snippet) so you can summarize, filter, or pass to other steps.

Output schema

  • results (array) — Each item typically has:
    • title — Page title.
    • url — Page URL.
    • snippet — Short excerpt from the page.

What it does: Searches for news articles.

Inputs

  • Query (required) — The news search query (e.g. “AI regulation 2024”).
  • Date range (optional) — Limit to articles from this period (e.g. last 7 days, last month).
  • Max results (optional) — Maximum number of articles to return.

Outputs

A list of news items with title, URL, source, and date for alerts or summaries.

Output schema

  • results / items (array) — Each item typically has:
    • title — Article headline.
    • url — Article URL.
    • source — Publication or source name.
    • date — Publication date (ISO or string).
    • snippet — Short excerpt (if available).

Scraping Webpages

What it does: Extracts data from a web page at a URL.

Inputs

  • URL (required) — The page URL to scrape (can be a variable). Must be publicly accessible or within allowed domains.
  • Extract images (optional) — When enabled, includes image URLs (or base64) in the output. Increases payload size.
  • Extract links (optional) — When enabled, includes a list of links found on the page.

Outputs

Page title, main text content, raw HTML, and optionally images and links for use in summaries or data extraction.

Output schema

  • title — Page title (from <title> or meta).
  • content — Main text content extracted from the page (plain text).
  • html — Raw HTML of the page (or main content area), if requested.
  • images (optional) — List of image URLs (or objects with url, alt).
  • links (optional) — List of link URLs (or objects with url, text).

Performing Web Research

What it does: Runs deeper web research (e.g. multi-query or research-style) and returns synthesized results. Use for due diligence, competitive research, or long-form answers. Inputs and outputs depend on the product (e.g. query, depth, max results; output may be a report or structured summary).

Browsing Agents

What it does: Uses a browsing agent to navigate web pages and extract or interact with content. Use when you need to follow links, fill forms, or get content that simple scrape doesn’t cover. Configuration and output schema are product-specific.

Search Unsplash Photos

What it does: Searches for stock photos from Unsplash.

Inputs

  • Query (required) — Search terms (e.g. “office meeting”, “nature landscape”). Can include variables.
  • Number of results (optional) — How many photos to return (e.g. 5, 10).
  • Orientation (optional) — Filter by aspect: landscape, portrait, or square. Affects dimensions of returned images.

Outputs

A list of photo objects with URLs and metadata for embedding in docs or emails.

Output schema

  • results (array) — Each item typically has:
    • id — Unsplash photo ID.
    • url — Full-size or regular image URL.
    • thumbnailUrl — Smaller thumbnail URL.
    • description — Photo description (if any).
    • photographer — Photographer name or credit.
    • width — Image width in pixels.
    • height — Image height in pixels.

Retrieving content from RSS Feeds

What it does: Fetches and processes RSS feed data.

Inputs

  • Feed URL(s) (required) — One or more RSS feed URLs (e.g. blog, news, calendar). Can be a variable with multiple URLs.
  • Max items (optional) — Maximum number of items to fetch per feed. Prevents huge payloads.
  • Filters (optional) — Optional filters (e.g. by date, keyword in title) if the product supports them.

Outputs

A list of feed items (title, link, description, date) for notifications or content pipelines.

Output schema

  • items (array) — Each item typically has:
    • title — Item title.
    • link — Item URL.
    • description — Summary or body (plain or HTML).
    • date / published — Publication date (ISO or string).
    • author (optional) — Author name or email.
    • categories (optional) — Tags or categories, if present in the feed.
Last updated on
AIEmail