.dinoprompts

The .dinoprompts file serves as your team's prompt library, allowing you to store and reuse battle-tested prompts that understand your data development patterns. This eliminates the need to recreate complex prompts and ensures consistency across your analytics engineering team.

Key Benefits

  • Centralized Library: Store tailored prompts for analytics engineering workflows

  • Time Savings: Access proven prompts instantly instead of recreating them

  • Team Knowledge Sharing: Distribute effective prompts across your organization

  • Context-Aware: Use variables for dynamic, situation-specific prompts

  • Fast Onboarding: New team members access best practices immediately


Step-by-Step Instructions

  1. Open DinoAI by clicking the DinoAI icon (πŸͺ„) in the left panel

  2. Access Prompt Creation by clicking the Prompt option in the chat input

    1. Alternative: Use the bracket symbol shortcut "[" to quickly find prompts

  3. Create the file by selecting "Add .dinoprompts" to automatically create a new file with built-in prompts

  4. Define your prompts using the YAML structure with name and prompt fields

  5. Access your prompts using the prompt quick-open feature in DinoAI


File Structure

The .dinoprompts file uses a simple YAML structure:

prompts:
  - name: "Prompt Name"
    prompt: |
      Multi-line prompt content
      Can include variables like {{ git.diff.withOriginDefaultBranch }}
      
  - name: "Another Prompt"
    prompt: "Single line prompts work too"

Available Variables

DinoAI provides built-in variables you can use in your prompts. Variables enable you to dynamically attach context to your prompts.

Vairable
Descriription

{{ git.diff.withOriginDefaultBranch }}

Includes the git diff between your current branch and the default branch of your repository

{{ editor.currentFile.path }}

Includes the file path of the current opened and selected file in your Code IDE

{{ editor.openFiles.path }}

Includes the file path of the all the opened files in your Code IDE


Best Practices

  1. Be Specific: Include clear requirements and expected outputs

  2. Use Variables: Leverage context-aware substitutions

  3. Provide Examples: Include sample inputs/outputs when helpful

  4. Iterate: Test and refine prompts based on results

  5. Descriptive Names: Use clear, purpose-driven prompt names

  6. Team Collaboration: Share and improve prompts regularly

Common Use Cases

Ready-made prompts for analytics engineering tasks:

  • Model Generation: Staging, intermediate, and mart models with proper structure

  • Code Review: Git diff analysis for best practices and performance

  • Documentation: Comprehensive model and column descriptions

  • Testing: Data quality tests for models and transformations

  • Optimization: Query performance improvements and bottleneck identification

  • Standards: Consistent formatting and naming convention enforcement

Last updated

Was this helpful?