The Knowledge Search action allows your AI assistant to perform semantic searches across your Knowledge Base and return the most relevant documents. This action is often used as a first step in workflows that require reference data to generate accurate answers or trigger next actions.

​
🔍 What It Does

This action queries your Knowledge Base using a search string (static or dynamic via parameters) and retrieves the most relevant documents based on similarity scoring, filters, tags, and categories.
⚙️ Use this action when your assistant needs to fetch context before replying or making decisions.

​
🖼️ Action Interface

Knowledge Search Action UI

​
⚙️ Configuration Options

Type: string The search term or phrase. You can reference a dynamic parameter using @, for example: @userQuery.Example: @userIssue, Reset password process
Type: string Comma-separated list of tags to include in the search.Example: billing, setup, FAQ
Type: string Comma-separated list of tags to exclude from the search.Example: internal, deprecated
Type: string Comma-separated list of Knowledge Base categories to search within.Example: Support, Sales
Type: number Maximum number of documents to return.
  • Set to 0 to return all matching documents.
Example: 5
Type: float Sets the similarity threshold (range: 0.0 to 1.0):
  • 0.0 = identical
  • 1.0 = return all
Example: 0.3 (returns closely matched documents)
Type: boolean When enabled, the search will also include content from tool outputs in addition to document bodies.
Type: boolean Limits the search to only the top matched category. This helps improve precision when multiple categories are available.

​
🪄 Tips

  • Use this action early in a workflow if the assistant needs knowledge context to proceed.
  • Pair with a Send Message or API Call action to utilize the retrieved documents.
  • Reference any defined parameters in your Query using @, e.g. @topic, @customerRequest.
Let me know if you’d like to see example tool chains using this action!
⌘I