Skip to main content

Prompt Engineering Tips

Prompt engineering is the art of crafting effective instructions for AI models like Roo Code. Well-written prompts lead to better results, fewer errors, and a more efficient workflow.

General Principles​

  • Be Clear and Specific: Clearly state what you want Roo Code to do. Avoid ambiguity.

    • Bad: Fix the code.
    • Good: Fix the bug in the calculateTotal function that causes it to return incorrect results.

  • Provide Context: Use Context Mentions to refer to specific files, folders, or problems.

    • Good: @/src/utils.ts Refactor the calculateTotal function to use async/await.

  • Break Down Tasks: Divide complex tasks into smaller, well-defined steps.

  • Give Examples: If you have a specific coding style or pattern in mind, provide examples.

  • Specify Output Format: If you need the output in a particular format (e.g., JSON, Markdown), specify it in the prompt.

  • Iterate: Don't be afraid to refine your prompt if the initial results aren't what you expect.

Thinking vs. Doing​

It's often helpful to guide Roo Code through a "think-then-do" process:

  1. Analyze: Ask Roo Code to analyze the current code, identify problems, or plan the approach.
  2. Plan: Have Roo Code outline the steps it will take to complete the task.
  3. Execute: Instruct Roo Code to implement the plan, one step at a time.
  4. Review: Carefully review the results of each step before proceeding.

Using Custom Instructions​

You can provide custom instructions to further tailor Roo Code's behavior. There are two types of custom instructions:

  • Global Custom Instructions: Apply to all modes.
  • Mode-Specific Custom Instructions: Apply only to a specific mode (e.g., Code, Architect, Ask, or a custom mode).

Custom instructions are added to the system prompt, providing persistent guidance to the AI model. You can use these to:

  • Enforce coding style guidelines.
  • Specify preferred libraries or frameworks.
  • Define project-specific conventions.
  • Adjust Roo Code's tone or personality.

See the Custom Instructions section for more details.

Handling Ambiguity​

If your request is ambiguous or lacks sufficient detail, Roo Code might:

  • Make Assumptions: It might proceed based on its best guess, which may not be what you intended.
  • Ask Follow-Up Questions: It might use the ask_followup_question tool to clarify your request.

It's generally better to provide clear and specific instructions from the start to avoid unnecessary back-and-forth.

Providing Feedback​

If Roo Code doesn't produce the desired results, you can provide feedback by:

  • Rejecting Actions: Click the "Reject" button when Roo Code proposes an action you don't want.
  • Providing Explanations: When rejecting, explain why you're rejecting the action. This helps Roo Code learn from its mistakes.
  • Rewording Your Request: Try rephrasing your initial task or providing more specific instructions.
  • Manually Correcting: If there are a few small issues, you can also directly modify the code before accepting the changes.

Examples​

Good Prompt:

@/src/components/Button.tsx Refactor the Button component to use the useState hook instead of the useReducer hook.

Bad Prompt:

Fix the button.

Good Prompt:

Create a new file named utils.py and add a function called calculate_average that takes a list of numbers and returns their average.

Bad Prompt:

Write some Python code.

Good Prompt:

@problems Address all errors and warnings in the current file.

Bad Prompt:

Fix everything.

By following these tips, you can write effective prompts that get the most out of Roo Code's capabilities.