Here's a comprehensive guide on how to construct effective SnapGPT prompts for pipeline generation: --- 🧠 How to Construct a SnapGPT Prompt for Pipeline Generation ✅ Core Principles (The 3 C's) | Principle | Description | |-----------|-------------| | Clear | Be precise so SnapGPT can determine the most suitable response | | Concise | Be brief to reduce response latency | | Consistent | Be consistent to avoid variability in results | --- 📐 Prompt Structure Best Practices A well-constructed prompt should include: 1. Action – What you want to do (e.g., read, write, transform, join) 2. Source System – Where the data comes from (e.g., Salesforce, AWS S3, REST API) 3. Target System – Where the data should go (e.g., Snowflake, Google BigQuery, a file) 4. Data/Object – What data is involved (e.g., Opportunities, CSV file, customer records) 5. Conditions (optional) – Any filtering, routing, or transformation logic --- 📈 Prompt Quality Examples
Good: Generate a pipeline to move data to S3.
Better: Generate a pipeline to move Salesforce data to an S3 bucket.
Best: Generate a pipeline to read Salesforce Opportunity records
and write them to an S3 bucket in JSON format.
--- 💡 Example Prompts That Work Well - Create a SnapLogic pipeline to join tables t1 and t2 and insert the output into table t3. - Read a file in CSV format and write to Google Sheets. - Get Salesforce opportunities and write to BigQuery. - Generate a pipeline with a JSON Generator that contains 10 sample Salesforce lead objects. - Get exchange rates from the European Central Bank and write the output to a JSON file. - Use an API to compute carbon emissions based on 50 kWh electricity consumption in FL, USA. - Download a CSV file from S3 and query it for records with zip code 78660. --- 🚫 What to Avoid - ❌ Vague prompts – e.g., "Generate a pipeline" (no context) - ❌ General/off-topic questions – e.g., "What is the capital of Florida?" - ❌ Sensitive data – Never include credentials, passwords, or PII in your prompt --- 🛠️ How to Use the Pipeline Generation Skill 1. Open SnapGPT from the Designer tab 2. Click Pipeline generation from the Featured Skills 3. Replace the example prompt with your specific use case 4. Use Extended Thinking mode for best accuracy in pipeline generation 5. Hit Return or click the submit button 6. Review the generated pipeline preview and choose to: - Replace the current pipeline, or - Import it into a new tab --- 📸 Pro Tip: Generate from an Image! You can also upload a flow diagram (.jpg, .jpeg, or .png, max 750 KB) to visually describe your pipeline. SnapGPT will interpret the diagram and generate the corresponding pipeline automatically. --- By following these guidelines, you'll get the most accurate and useful pipeline generations from SnapGPT! 🚀
I understand you're working on creating a finalized output for Box AI integration, and you've identified that DOCX and MD formats work well for visual interpretation while JSON doesn't meet your needs for this use case. To help you build the right solution, I'd like to understand more about your specific requirements: 1. What type of data are you starting with? (e.g., structured data from a database, text content, reports, etc.) 2. Which output format would you prefer - DOCX or Markdown (MD)? 3. What's the source of your data? (e.g., database, API, files, etc.) 4. Do you want me to create the pipeline for you, or are you looking for guidance on how to approach this? Once I understand these details, I can either: - Create a complete pipeline that transforms your data into the appropriate format and uploads it to Box - Provide guidance on the best Snaps and approach to use for your specific scenario Both DOCX and MD are excellent choices for Box AI since they preserve formatting and structure that AI can interpret visually, unlike JSON which is purely data-focused. Let me know your preferences and I'll help you build the right solution!
I'll search for information about generating DOCX and Markdown files in SnapLogic to provide you with the available options. Yes, there are several ways to generate DOCX files and Markdown files in SnapLogic: For DOCX Files 1. Google Drive Read Snap The Google Drive Read Snap can export Google Docs directly to DOCX format: - Set the Export MIME Type to application/vnd.openxmlformats-officedocument.wordprocessingml.document - Or use "Default" which automatically exports Google Docs as DOCX files - This downloads the file as binary content that you can then save using a File Writer Snap 2. XML Generator Snap You can generate DOCX files using the XML Generator Snap since DOCX files are essentially XML-based: - Create an XML template that follows the Office Open XML format - Use Apache Velocity templating to insert dynamic content - Convert the output to binary format using DocumentToBinary Snap - Save the file with a .docx extension For Markdown Files 1. HTML to Markdown Converter Snap This specialized Snap converts HTML content directly to Markdown: - Input: HTML content (as binary file or document string) - Output: Markdown content (as binary or document format) - Perfect for converting web content or HTML documents to Markdown 2. Text-based Generation You can generate Markdown files using standard text processing: - Use JSON Generator or Mapper Snaps to create Markdown-formatted text - Apply Markdown syntax (headers with , lists with -, etc.) - Save the output as a .md file using File Writer Snap Example Workflow for Markdown Generation 1. HTTP Client → Fetch HTML content from a web source 2. HTML to Markdown Converter → Convert HTML to Markdown format 3. File Writer → Save as .md file Example Workflow for DOCX Generation 1. Data Source → Your input data 2. XML Generator → Create DOCX-compatible XML structure 3. DocumentToBinary → Convert to binary format 4. File Writer → Save as .docx file Both approaches allow for dynamic content generation based on your pipeline data, making them suitable for automated document creation workflows.