Using Knowledge Base in Workflows

Overview

Knowledge Bases can be integrated into workflows in two ways:

Content ingestion: Add, update, or delete content sources dynamically from workflows
AI agent queries: Enable Custom Agent nodes to query Knowledge Bases for contextual information

Content ingestion from workflows

You can manage Knowledge Base content sources directly from workflows using Knowledge Base operations. This allows you to dynamically update Knowledge Bases based on process data.

Adding knowledge base operations

Open your workflow

Navigate to the workflow where you want to add a Knowledge Base operation

Add a task node

Add a new task node or select an existing one

Add Knowledge Base action

Add a new action and select Knowledge Base Operation as the action type

Configure the operation

Fill in the required configuration fields

Configuration fields

Knowledge Base

select

required

Select a Knowledge Base from:

Local project libraries
Dependency libraries

Operation

select

required

Choose the operation to perform on the content source

Content Source

string

required

The name of the content source to operate on. You can:

Select an existing content source from the dropdown
Type a new content source name
Use a workflow variable (e.g., ${contentSourceName})
Use a configuration parameter

Content

string

required

The JSON content to ingest. Typically a workflow variable containing the data to be indexed.

Available operations

The following operations are available for managing content sources from workflows:

Append content

Add new content to a content source without removing existing chunks.

Property	Behavior
Creates if missing?	Yes
Deletes existing content?	No

Use case: continuously adding new information to a Knowledge Base, such as adding new product updates or FAQ entries. Example:

{
  "knowledge_base": "Product Documentation KB",
  "operation": "Append Content",
  "content_source": "${process.contentSource}",
  "content": "${process.productUpdate}"
}

Replace content

Replace all existing content in a content source with new content.

Property	Behavior
Creates if missing?	Yes
Deletes existing content?	Yes (content only)

Use case: updating documentation to a new version where the previous version should be completely replaced. Example:

{
  "knowledge_base": "Policy Documents KB",
  "operation": "Replace Content",
  "content_source": "Employee Handbook",
  "content": "${process.updatedHandbook}"
}

Replace Content permanently deletes all existing chunks from the content source before adding new content.

Delete content source

Remove a content source and all its associated chunks completely.

Property	Behavior
Creates if missing?	No
Deletes existing content?	Yes (object + content)

Use case: removing outdated or deprecated content that should no longer be available to AI agents. Example:

{
  "knowledge_base": "Temporary KB",
  "operation": "Delete Content Source",
  "content_source": "Deprecated Policy"
}

Content format

Content ingested from workflows must be in JSON format. The JSON can contain:

Plain text strings
Structured data objects
Arrays of information
Any JSON-serializable data

Example content:

{
  "title": "Product Update - Q4 2025",
  "sections": [
    {
      "heading": "New Features",
      "content": "We've introduced three major features this quarter..."
    },
    {
      "heading": "Bug Fixes",
      "content": "Resolved critical issues affecting performance..."
    }
  ],
  "metadata": {
    "version": "2.5.0",
    "release_date": "2025-12-15"
  }
}

Using variables and parameters

You can use workflow variables and configuration parameters in Knowledge Base operations:

Workflow Variables
Configuration Parameters

Reference data from the current process instance:

{
  "content_source": "${application.documentType}",
  "content": "${application.documentContent}"
}

Use environment-specific values:

{
  "knowledge_base": "${config.defaultKnowledgeBase}",
  "content_source": "${config.contentSourcePrefix}_${process.id}"
}

Console logs

Knowledge Base operations generate detailed console logs showing the execution status:

The console log displays:

Operation

string

The Knowledge Base operation that was executed

Knowledge Base

string

The name of the Knowledge Base

Content Source

string

The name of the content source that was modified

Status

string

Success or error status of the operation

Chunks Created

number

Number of chunks created (for Append and Replace operations)

Chunks Deleted

number

Number of chunks deleted (for Replace and Delete operations)

Error Message

string

Detailed error message if the operation failed

Using knowledge bases in custom agent nodes

Custom Agent nodes can query Knowledge Bases to retrieve contextual information when generating responses.

Configuring custom agent with knowledge base

Add Custom Agent node

Add a Custom Agent node to your workflow

Configure the agent

Set up the basic agent configuration (model, prompt, etc.)

Enable Knowledge Base

In the Knowledge Base section, enable the Knowledge Base feature

Select Knowledge Base

Choose a Knowledge Base from the available options

Configure query parameters

Set the query parameters for retrieving chunks

Knowledge base query configuration

Knowledge Base

select

required

Select the Knowledge Base that the agent should query

Content Source Filter

select

Options:

All content sources (default)
Specific content source(s)

Filter which content sources the agent should search

Max. Number of Results

number

Maximum number of chunks to retrieve (1-10). Default: 5

Min. Relevance Score

percentage

Minimum relevance score threshold (0-100%). Only chunks above this score will be included. Default: 70%

Metadata Filters

object

Optional metadata filters to apply (exact match with AND logic)

Metadata filtering is a low-priority feature for the first iteration.

How the agent uses knowledge base

When a Custom Agent node executes with Knowledge Base enabled:

Query generation: The agent analyzes the user’s input and generates a search query
Chunk retrieval: The Knowledge Base is queried using semantic search
Relevance filtering: Only chunks meeting the minimum relevance score are returned
Context inclusion: Retrieved chunks are added to the agent’s context
Response generation: The agent generates a response using the retrieved information

ReAct model with knowledge base

When using the ReAct (Reasoning and Acting) model, the agent can:

Reason: Determine if Knowledge Base information is needed
Act: Query the Knowledge Base as a tool
Iterate: Make multiple queries if needed for comprehensive answers

Example ReAct flow:

User: "What are the new features in version 2.5?"

Agent Reasoning: I need to search the Knowledge Base for version 2.5 information

Agent Action: Query Knowledge Base with "new features version 2.5"

Knowledge Base: Returns 3 relevant chunks about version 2.5 features

Agent Reasoning: I have enough information to answer the question

Agent Response: "Version 2.5 introduces three major features..."

Console logs for knowledge base queries

When an agent queries a Knowledge Base, the console log shows detailed information:

The console log includes:

Query

string

The search query generated by the agent

Content Sources Searched

array

List of content sources that were searched

Chunks Retrieved

number

Total number of chunks retrieved

Chunks Used

array

Details of each chunk used, including:

Chunk content
Relevance score
Content source
Metadata

Execution Time

duration

Time taken to retrieve and process chunks

Use console logs to understand:

Which chunks are most relevant for different queries
Whether relevance score thresholds need adjustment
If content sources are providing useful information

Example workflows

Dynamic documentation updates

Update product documentation whenever a new release is deployed:

Workflow: Product Release
Nodes:
  1. Receive Release Data
  2. Knowledge Base Operation
     - KB: "Product Documentation KB"
     - Operation: "Replace Content"
     - Content Source: "Release Notes ${release.version}"
     - Content: "${release.documentation}"
  3. Notify Support Team

Customer support with contextual AI

Use Knowledge Base in a customer support chatbot:

Workflow: Customer Support Chat
Nodes:
  1. Receive Customer Message
  2. Custom Agent Node
     - Model: GPT-4
     - Knowledge Base: "Support Documentation KB"
     - Content Sources: All
     - Min. Relevance: 75%
     - Max Results: 5
  3. Send Response to Customer

Multi-source knowledge ingestion

Aggregate information from multiple systems:

Workflow: Weekly Knowledge Update
Nodes:
  1. Fetch from CRM System
  2. Append to KB
     - Content Source: "CRM Updates"
  3. Fetch from Documentation Site
  4. Append to KB
     - Content Source: "Documentation"
  5. Fetch from Support Tickets
  6. Append to KB
     - Content Source: "Support Insights"

Error handling

Common errors and solutions

Content source not found

Error: Content source "Example" does not exist in Knowledge Base "KB Name"Cause: Attempting to append or replace content in a non-existent content sourceSolution:

Verify the content source name is correct
Use “Append Content” which creates the content source if it doesn’t exist
Check for typos in content source names

Invalid JSON content

Error: Failed to parse JSON contentCause: The content provided is not valid JSONSolution:

Ensure workflow variables contain valid JSON
Validate JSON structure before passing to Knowledge Base operation
Check for special characters that may break JSON formatting

Knowledge Base not accessible

Error: Knowledge Base "KB Name" not found or not accessibleCause: The Knowledge Base doesn’t exist or you don’t have permissionsSolution:

Verify the Knowledge Base exists in the project or dependencies
Check access permissions for the Knowledge Base
Ensure the Knowledge Base is published if using from dependencies

Chunking failed

Error: Failed to process content into chunksCause: Error occurred while creating chunks from the contentSolution:

Check console logs for detailed error messages
Verify the content is in the expected format
Retry the operation
Contact support if the issue persists

Best practices

Content ingestion

Workflow content ingestion tips:

Use meaningful content source names that include versioning
Include metadata in your JSON content for better organization
Test operations in isolation before adding to production workflows
Handle errors gracefully with try-catch patterns in your workflow

Agent configuration

Custom Agent with Knowledge Base:

Start with a higher minimum relevance score (70-80%) to avoid irrelevant results
Adjust max results based on your use case (fewer for focused answers, more for comprehensive coverage)
Test different queries to understand chunk retrieval behavior
Monitor console logs to optimize configuration

Performance

Optimize performance:

Keep content sources focused and organized by topic
Use content source filters when you know which source contains relevant information
Avoid ingesting extremely large documents without proper structuring
Regular cleanup of deprecated content sources

Limitations

Current limitations:

Only one Knowledge Base per Custom Agent node
Cannot ingest documents directly from Document Plugin
Metadata filters use exact match with AND logic only (first iteration)
Custom metadata cannot be defined

Next steps

Testing Operations

Learn how to test Knowledge Base operations before production use

Managing Content

Understand content source management in detail

Custom Agent Nodes

Complete guide to Custom Agent nodes

Workflow Data Models

Understanding workflow variables and data

Getting Started

Platform overview

FlowX.AI Designer

Building projects

Building blocks

Platform deep dive

API reference

​Overview

​Content ingestion from workflows

​Adding knowledge base operations

​Configuration fields

​Available operations

​Append content

​Replace content

​Delete content source

​Content format

​Using variables and parameters

​Console logs

​Using knowledge bases in custom agent nodes

​Configuring custom agent with knowledge base

​Knowledge base query configuration

​How the agent uses knowledge base

​ReAct model with knowledge base

​Console logs for knowledge base queries

​Example workflows

​Dynamic documentation updates

​Customer support with contextual AI

​Multi-source knowledge ingestion

​Error handling

​Common errors and solutions

​Best practices

​Content ingestion

​Agent configuration

​Performance

​Limitations

​Next steps

Testing Operations

Managing Content

​Related resources

Custom Agent Nodes

Workflow Data Models

Overview

Content ingestion from workflows

Adding knowledge base operations

Configuration fields

Available operations

Append content

Replace content

Delete content source

Content format

Using variables and parameters

Console logs

Using knowledge bases in custom agent nodes

Configuring custom agent with knowledge base

Knowledge base query configuration

How the agent uses knowledge base

ReAct model with knowledge base

Console logs for knowledge base queries

Example workflows

Dynamic documentation updates

Customer support with contextual AI

Multi-source knowledge ingestion

Error handling

Common errors and solutions

Best practices

Content ingestion

Agent configuration

Performance

Limitations

Next steps

Related resources