Tools Agent

The Tools Agent is like having an AI assistant that can actually do things. Give it a task like “research competitor pricing” and it will automatically figure out which tools to use, visit websites, extract information, and compile results - all on its own.

This is true AI automation that can adapt and make decisions as it works toward your goal.

How it works

You describe what you want accomplished, and the AI agent creates a plan, selects the right tools for each step, executes the plan, and adapts if something doesn’t work as expected.

graph LR
  Goal[Your Goal] --> Plan[AI Planning]
  Plan --> Tools[Select Tools]
  Tools --> Execute[Execute Steps]
  Execute --> Adapt[Adapt & Continue]
  Adapt --> Result[Final Results]
  style Plan fill:#6d28d9,stroke:#fff,color:#fff

Setup guide

1. Describe Your Task: Write a clear description of what you want the agent to accomplish.

2. Choose Available Tools: Select which tools the agent can use to complete the task.

3. Set Limits: Configure maximum steps and error handling to keep the agent focused.

4. Let It Work: The agent will automatically plan, execute, and adapt until the task is complete.

Practical example: Competitive research

Let’s create an agent that automatically researches competitor pricing across multiple websites.

Task 1: Research Prices

• Instruction: “Visit 3 competitor websites and extract their pricing information.”
• Tools Given: Get All Text, AI Text Processor, Field Editor.
• Limit: 8 steps maximum.

Task 2: Lead Collection

• Instruction: “Extract contact information from tech company websites.”
• Tools Given: Get HTML, Field Editor, Filter.
• Limit: 6 steps maximum.

Task 3: Analyze Reviews

• Instruction: “Analyze product reviews and summarize key insights.”
• Tools Given: Get All Text, AI Text Processor, Questions & Answers.
• Limit: 10 steps maximum.

What makes it smart

Traditional Automation	Tools Agent
Fixed sequence of steps	Adapts plan based on results
Breaks if something changes	Tries alternative approaches
Requires manual tool selection	Automatically chooses best tools
No error recovery	Learns from failures and adjusts

Configuration settings

Setting	Purpose	Recommended Values
Task Description	What you want accomplished	Be specific about the end goal
Available Tools	Tools the agent can use	Only include relevant tools
Max Iterations	Maximum steps allowed	5-10 for most tasks
Planning Mode	How to approach the task	”adaptive” for flexibility

Start with clear goals

Instead of “research competitors,” try “visit 3 competitor websites and extract pricing for their basic and premium plans.” Specific goals get better results.

Real-world examples

Market research automation

Automatically gather competitive intelligence:

Task: "Research pricing and features from 5 SaaS competitors"
Tools: GetAllTextFromLink, BasicLLMChain, EditFields
Max Steps: 12
Result: Comprehensive competitive analysis report

Lead generation

Find and qualify potential customers:

Task: "Find contact information for tech startups in San Francisco"
Tools: GetHTMLFromLink, EditFields, Filter
Max Steps: 8
Result: Qualified lead list with contact details

Content monitoring

Track mentions and sentiment across websites:

Task: "Monitor 10 news sites for mentions of our company"
Tools: GetAllTextFromLink, BasicLLMChain, QANode
Max Steps: 15
Result: Sentiment analysis and mention tracking

Troubleshooting

• Agent gets stuck in loops: Make your task description more specific with clear success criteria, or reduce max iterations.
• Wrong tool selection: Limit available tools to only those relevant for your specific task.
• Incomplete results: Increase max iterations or break complex tasks into smaller, focused goals.
• Slow execution: Reduce the number of available tools or optimize the tools being used for better performance.