What it does
Performs a RAG (Retrieval-Augmented Generation) query to retrieve relevant information from your training materials and knowledge base, then adds it to the agent's context for answering.
When to use it
When you want the agent to reference specific documentation, product information, or knowledge base articles in its response.
Can combine with
@ask_question
@show_options
@mention_specifically
Configuration
None required—uses your existing training materials automatically.
Syntax
@recommend [context or topic description]
Examples
Basic product recommendation:
1. @ask_question: "What product are you interested in?" @show_options["Product A", "Product B", "Product C"] 2. @recommend information about the selected product 3. @ask_question: "Does this sound like what you need?" @show_options["Yes", "No"]
Solution recommendation:
1. @ask_question: "What issue are you experiencing?" 2. @recommend a solution for the issue. @show_options["Issue resolved", "Still need help"]
Specific topic:
1. @recommend pricing information for Enterprise plans 2. @ask_question: "Would you like to discuss a custom quote?" @show_options["Yes", "No"]
Tips
- Provide context in the recommendation instruction for better results
- Example:
@recommend pricing information for Enterprise plansis better than just@recommend
- Combine with questions to create consultative flows
- Works best when you have comprehensive training materials uploaded
- The agent will intelligently select relevant information from your knowledge base
How It Works
- Agent analyzes the visitor's context and your instruction
- Searches your training materials for relevant content
- Retrieves the most relevant information
- Incorporates it into the response naturally
- Combines with other tools in the same step if specified
Learn More
- Getting Started with Playbooks - Overview and all tools
- Core Concepts & Structure - Understanding playbook structure
- Complete Playbook Examples - See RAG in action
- Best Practices & Tips - Expert guidance