Skip to main content

Frequently Asked Questions

This comprehensive FAQ addresses common questions about Thunderflow to help you get started and make the most of its capabilities.

General​

What is Thunderflow?​

Thunderflow is an AI-powered autonomous coding agent that integrates directly into your editor, providing intelligent assistance for your development workflow.

How does Thunderflow work?​

Thunderflow leverages advanced large language models (LLMs) to understand your requests and convert them into actionable operations. It can:

  • Read and write files within your project
  • Execute commands in your VS Code terminal
  • Perform web browsing (when enabled)
  • Utilize external tools through the Model Context Protocol (MCP)

You interact with Thunderflow through an intuitive chat interface where you provide instructions and review/approve its proposed actions.

What capabilities does Thunderflow offer?​

Thunderflow assists with a wide range of development tasks, including:

  • Code generation from natural language descriptions
  • Code refactoring and optimization
  • Bug identification and resolution
  • Documentation creation and maintenance
  • Code explanation and analysis
  • Codebase-specific question answering
  • Automation of repetitive tasks
  • Project and file creation

Is Thunderflow free to use?​

The Thunderflow extension itself is free and open-source. However, Thunderflow relies on external API providers for its AI capabilities, including Anthropic, OpenAI, OpenRouter, and Requesty. These providers typically charge for API usage based on token consumption. You'll need to create an account and obtain an API key from your preferred provider. For detailed setup instructions, see Setting Up Your First AI Provider.

What should I be aware of when using Thunderflow?​

While Thunderflow is a powerful development tool, responsible usage is important:

  • Verification is essential - Always carefully review Thunderflow's proposed changes before approval
  • Command execution requires caution - Be vigilant when allowing Thunderflow to run commands, especially with auto-approval enabled
  • Internet access considerations - If using a provider with web browsing capabilities, be mindful that Thunderflow could potentially access sensitive information

Setup & Installation​

How do I install Thunderflow?​

For comprehensive installation instructions, refer to the Installation Guide.

Which AI providers does Thunderflow support?​

Thunderflow integrates with numerous AI providers, including:

How do I obtain an API key?​

Each AI provider has its own process for API key acquisition. For provider-specific instructions, see Setting Up Your First AI Provider.

Can I use Thunderflow with locally hosted models?​

Yes, Thunderflow supports local model execution through Ollama and LM Studio. For implementation details, see Using Local Models.

Usage​

How do I initiate a new task?​

Open the Thunderflow panel () and enter your task in the chat interface. Be specific and clear about your requirements. For guidance on effective request formulation, see Typing Your Requests.

What are Thunderflow modes?​

Modes represent specialized personas that Thunderflow can adopt, each with distinct capabilities and focus areas:

  • Code: For general-purpose coding and development tasks
  • Architect: For technical planning, architecture, and leadership
  • Ask: For information retrieval and question answering
  • Debug: For systematic problem identification and resolution

You can also create Custom Modes tailored to your specific needs.

How do I switch between different modes?​

Use the dropdown menu in the chat input area to select your desired mode, or use the / command to switch directly to a specific mode.

How do tools function in Thunderflow?​

Tools are Thunderflow's interface for system interaction. Thunderflow automatically selects and employs appropriate tools to complete your tasks without requiring direct tool invocation. You'll be prompted to approve or reject each tool use before execution.

What are context mentions?​

Context mentions provide Thunderflow with specific project information such as files, folders, or problems. Use the "@" symbol followed by the item you want to reference (e.g., @src/file.ts, @problems).

Does Thunderflow have internet access capabilities?​

Yes, if you're using a provider with a model that supports web browsing. Consider the security implications when enabling this feature.

Can Thunderflow execute terminal commands?​

Yes, Thunderflow can run commands in your VS Code terminal. You'll be prompted to approve each command before execution, unless you've enabled auto-approval. Exercise caution with command auto-approval. For troubleshooting terminal command issues, consult the Shell Integration Guide.

How can I provide feedback to Thunderflow?​

You can provide feedback by approving or rejecting Thunderflow's proposed actions and by using the feedback field for additional comments.

What customization options does Thunderflow offer?​

Thunderflow provides several customization pathways:

  • Custom Instructions: Configure general or mode-specific instructions
  • Custom Modes: Create personalized modes with tailored prompts and tool permissions
  • .thunderflowrules Files: Implement project-specific guidelines through .thunderflowrules files
  • Settings: Adjust various configuration options including auto-approval, diff editing, and more

Does Thunderflow support automatic action approval?​

Yes, Thunderflow includes settings that can automatically approve certain actions. Learn more about these options here.

Advanced Features​

Can Thunderflow operate offline?​

Yes, when using a local model.

What is the Model Context Protocol (MCP)?​

MCP is a protocol enabling Thunderflow to communicate with external servers, extending its functionality through custom tools and resources.

Can I develop custom MCP servers?​

Yes, you can create custom MCP servers to enhance Thunderflow with specialized functionality. For implementation details, refer to the MCP documentation.

Troubleshooting​

What should I do if Thunderflow becomes unresponsive?​

  • Verify your API key is valid and hasn't expired
  • Check your internet connection
  • Confirm the operational status of your chosen API provider
  • Try restarting VS Code
  • If issues persist, report the problem to support@thunderflow.ai

How should I interpret error messages?​

Error messages typically provide information about the underlying issue. If you're uncertain about resolution steps, seek assistance through community forums.

How can I undo unwanted changes made by Thunderflow?​

Thunderflow utilizes VS Code's built-in editing capabilities. Use the standard "Undo" command (Ctrl/Cmd + Z) to revert changes. Additionally, if experimental checkpoints are enabled, Thunderflow can revert file modifications.

Where can I report bugs or suggest features?​

Please submit bug reports or feature suggestions to support@thunderflow.ai.

Is this documentation incorrect or incomplete? Report an issue via Email