Understanding Thunderflow AI Tools

Thunderflow AI leverages a sophisticated set of tools to interact with your codebase and development environment. These specialized assistants perform targeted actions such as reading files, implementing code changes, executing commands, or searching through your project. Tools provide seamless automation for common development tasks without requiring manual intervention.

The Tool Execution Workflow

Simply describe your objective in natural language, and Thunderflow AI will orchestrate the process:

1. Intelligently select the most appropriate tool based on your specific request
2. Present the tool with its configured parameters for your review and approval
3. Execute the approved tool and display the results for your evaluation
4. Continue this iterative process until your task is fully completed

Tool Categories and Capabilities

Category	Purpose	Tool Names
Read	Access and analyze file content and code structure	read_file, search_files, list_files, list_code_definition_names
Edit	Create new files or modify existing code with precision	write_to_file, apply_diff
Execute	Run system commands and perform operational tasks	execute_command
Browser	Interact with and manipulate web content	browser_action
Workflow	Manage task progression and contextual understanding	ask_followup_question, attempt_completion, switch_mode, new_task

Tool Interaction Example

Here's an illustration of a typical tool interaction sequence:

The tool approval interface displays Save/Reject options and Auto-approve settings for your control.

User: Create a file named greeting.js that logs a greeting message

Thunderflow AI: (Proposes the write_to_file tool as shown in the image above)

<write_to_file>
<path>greeting.js</path>
<content>
function greet(name) {
  console.log(`Hello, ${name}!`);
}

greet('World');
</content>
<line_count>5</line_count>
</write_to_file>

User: (Clicks "Save" in the interface)

Thunderflow AI: (Confirms file creation)

Tool Safety and Approval Mechanisms

Every tool operation requires your explicit authorization. When Thunderflow AI proposes a tool, the interface presents:

• A "Save" button to approve and execute the proposed action
• A "Reject" button to decline the suggestion and provide alternative direction
• An optional "Auto-approve" setting for streamlining trusted, repetitive operations

This comprehensive safety framework ensures you maintain complete control over which files are modified, what commands are executed, and how your codebase evolves. Always carefully review tool proposals before authorizing their execution.

Core Tools Reference

Tool Name	Description	Category
read_file	Retrieves and displays file content with line numbering for reference	Read
search_files	Performs advanced text or pattern searches across multiple files	Read
list_files	Provides directory structure and file listings for specified locations	Read
list_code_definition_names	Extracts and displays code structure elements like classes and functions	Read
write_to_file	Creates new files or completely replaces existing file content	Edit
apply_diff	Implements precise, targeted modifications to specific sections of a file	Edit
execute_command	Executes system commands within the VS Code terminal environment	Execute
browser_action	Performs automated interactions with web content in a headless browser	Browser
ask_followup_question	Requests additional clarification when needed to proceed	Workflow
attempt_completion	Signals task completion and presents final results	Workflow
switch_mode	Transitions to a different operational mode for specialized tasks	Workflow
new_task	Initiates a new subtask with specific starting parameters	Workflow

Learn More About Tools

For comprehensive information about each tool, including detailed parameter references and advanced usage strategies, refer to the Tool Use Overview documentation.