Skip to main content

Using AI Assistant Slash Commands

This guide explains how to use Coginiti's AI Assistant slash commands, which provide quick access to web content, real-time searches, and enhanced AI capabilities directly within your data analysis workflow.

Overview​

AI Assistant slash commands are specialized shortcuts that extend the AI Assistant's capabilities beyond code generation and query optimization. These commands enable you to fetch web content, perform real-time searches, and access additional AI features without leaving your analysis environment.

Available Slash Commands​

  • /fetch [url] - Load and analyze content from web pages
  • /web [query] - Perform real-time web searches for current information
  • /help - Display list of available commands and usage instructions

Key Benefits​

Real-time Information Access:

  • Get current data and information from the web
  • Access external documentation and resources
  • Incorporate live data into your analysis workflow

Enhanced Analysis Context:

  • Load external datasets and documentation
  • Research industry trends and benchmarks
  • Access API documentation and technical references

Streamlined Workflow:

  • No need to switch between applications
  • Direct integration with AI-powered analysis
  • Immediate access to search results and web content

Usage Limits and Guidelines​

Rate Limits​

To ensure optimal performance and prevent misuse, slash commands have the following usage limits:

Per User Limits:

  • 10 requests per hour - Hourly limit for all slash command usage
  • 100 requests per day - Daily limit across all slash commands
  • Limits reset at the beginning of each hour/day

Fair Usage:

  • Limits apply to all users individually
  • Commands count toward limits when executed
  • Failed requests may still count toward limits
Managing Usage Limits

Plan your slash command usage strategically. Use /web for quick searches and /fetch for detailed content analysis. Monitor your usage to stay within limits for critical analysis work.

Best Practices​

Efficient Usage:

  • Combine multiple queries when possible
  • Use specific, targeted searches rather than broad queries
  • Cache important information locally when retrieved

Strategic Planning:

  • Prioritize essential searches and content fetching
  • Use commands during critical analysis phases
  • Plan research sessions within rate limit windows

/fetch Command​

The /fetch command loads and analyzes content from web pages, making external information available for AI analysis and integration with your data work.

Syntax​

/fetch [url]

Parameters​

  • url - The complete URL of the web page to fetch (including https://)

Use Cases​

API Documentation Access​

/fetch https://docs.snowflake.com/en/sql-reference/sql/create-table

Result: Load Snowflake CREATE TABLE documentation for reference while building data models.

Dataset Documentation​

/fetch https://github.com/user/dataset/blob/main/README.md

Result: Access dataset documentation to understand data structure and column definitions.

Industry Reports and Analysis​

/fetch https://example.com/industry-report-2024.html

Result: Load external reports for comparative analysis and benchmarking.

Technical Specifications​

/fetch https://parquet.apache.org/docs/file-format/

Result: Access file format specifications for data processing decisions.

How /fetch Works​

Content Processing​

  1. URL Validation - Ensures the provided URL is accessible and valid
  2. Content Retrieval - Fetches the web page content securely
  3. Content Parsing - Extracts relevant text and structured information
  4. AI Integration - Makes content available for AI analysis and questions

Content Types Supported​

  • HTML Pages - Standard web pages with text and structured content
  • Documentation Sites - Technical documentation and API references
  • Markdown Files - GitHub README files and documentation
  • Plain Text - Simple text files and content

Content Limitations​

  • Large Files - Very large pages may be truncated or summarized
  • Dynamic Content - JavaScript-generated content may not be fully captured
  • Protected Content - Password-protected or restricted content cannot be accessed
  • Media Files - Images, videos, and binary files are not processed

Working with Fetched Content​

Immediate Analysis​

After fetching content, you can immediately ask questions about it:

User: /fetch https://docs.databricks.com/sql/language-manual/sql-ref-syntax-qry-select.html

AI Assistant: [Loads Databricks SELECT documentation]

User: How does Databricks handle window functions differently from standard SQL?

AI Assistant: Based on the documentation I just fetched, Databricks supports standard window functions but has these key differences: [detailed analysis of fetched content]

Integration with Query Development​

User: /fetch https://spark.apache.org/docs/latest/sql-ref-functions-builtin.html

AI Assistant: [Loads Spark function reference]

User: Help me write a query using Spark's array functions to process JSON data

AI Assistant: Using the Spark functions I just loaded, here's how to process JSON arrays: [code generation using fetched documentation]

Comparative Analysis​

User: /fetch https://docs.aws.amazon.com/redshift/latest/dg/r_CREATE_TABLE_AS.html

AI Assistant: [Loads Redshift CREATE TABLE AS documentation]

User: Compare this with Snowflake's CREATE TABLE AS syntax from our database metadata

AI Assistant: Comparing the Redshift documentation with your Snowflake setup: [detailed comparison using both sources]

/web Command​

The /web command performs real-time web searches to find current information, trends, and answers to specific questions.

Syntax​

/web [query]

Parameters​

  • query - The search terms or question you want to research

Use Cases​

/web latest data visualization trends 2024

Result: Find current trends in data visualization for modernizing reports and dashboards.

Technical Problem Solving​

/web PostgreSQL connection timeout solutions

Result: Research solutions for database connectivity issues you're experiencing.

Competitive Intelligence​

/web market share business intelligence tools 2024

Result: Access current market data for strategic planning and tool selection.

Best Practices Research​

/web SQL query optimization best practices large datasets

Result: Find current best practices for optimizing queries on large datasets.

Search Strategy Tips​

Effective Query Construction​

Specific Terms:

  • Use specific technical terms and product names
  • Include version numbers when relevant
  • Add timeframe qualifiers (2024, latest, recent)

Good Examples:

/web Snowflake cost optimization strategies 2024
/web Apache Spark memory management best practices
/web Python pandas performance tuning large datasets

Less Effective:

/web database optimization
/web data processing
/web analytics tools

Research Workflows​

Problem-Solution Research:

  1. Identify specific issue - Define the exact problem you're facing
  2. Search for solutions - Use /web to find current solutions and approaches
  3. Apply findings - Integrate research results with your AI-powered analysis
  4. Validate approach - Use AI Assistant to help evaluate and implement solutions

Trend Analysis:

  1. Market research - Search for industry trends and developments
  2. Competitive analysis - Find information about competitor approaches
  3. Technology updates - Research new features and capabilities
  4. Strategic planning - Use findings for project and technology decisions

Working with Search Results​

Immediate Follow-up Questions​

User: /web latest Apache Iceberg performance benchmarks

AI Assistant: [Returns search results about Iceberg performance]

User: Based on these results, should we consider migrating from Delta Lake to Iceberg?

AI Assistant: Based on the performance benchmarks I found, here's my analysis of Delta Lake vs Iceberg for your use case: [detailed comparison using search results]

Integration with Current Work​

User: /web Tableau alternatives open source 2024

AI Assistant: [Returns search results about open source visualization tools]

User: Which of these would work best with our current PostgreSQL and Snowflake setup?

AI Assistant: Considering your PostgreSQL and Snowflake environment, here are the best options from my search results: [tailored recommendations]

/help Command​

The /help command displays comprehensive information about available AI Assistant features and commands.

Syntax​

/help

What /help Provides​

Command Reference​

  • Complete list of all available slash commands
  • Syntax examples for each command
  • Usage guidelines and best practices
  • Rate limit information and current usage status

Feature Overview​

  • AI Assistant capabilities beyond slash commands
  • Integration features with database connections and metadata
  • Keyboard shortcuts for AI features
  • Tips and tricks for effective AI interaction

Getting Started Guidance​

  • Quick start tips for new users
  • Common use cases and example workflows
  • Troubleshooting common issues
  • Links to additional documentation

When to Use /help​

New User Orientation​

  • First time using AI Assistant slash commands
  • Learning available features and capabilities
  • Understanding usage limits and guidelines

Feature Discovery​

  • Exploring new capabilities after updates
  • Finding optimal workflows for your use cases
  • Discovering advanced features you might have missed

Troubleshooting​

  • Command syntax issues or errors
  • Rate limiting questions and current status
  • Feature availability questions

Advanced Usage Patterns​

Combining Commands for Research Workflows​

Comprehensive Research Process​

  1. Start with broad search using /web to identify relevant resources
  2. Fetch specific documentation using /fetch for detailed information
  3. Ask targeted questions about the fetched content
  4. Apply insights to your current analysis or development work

Example Research Workflow​

Step 1: /web modern data warehouse architecture best practices 2024

Step 2: /fetch https://docs.aws.amazon.com/wellarchitected/latest/analytics-lens/well-architected-analytics-lens.html

Step 3: Based on the search results and AWS documentation, how should we structure our data warehouse migration from PostgreSQL to Snowflake?

Step 4: Create a CoginitiScript to implement the recommended data pipeline structure

Integration with Database Work​

Documentation-Driven Development​

User: /fetch https://docs.databricks.com/delta/delta-batch.html

AI Assistant: [Loads Delta Lake documentation]

User: Using this documentation and our current Snowflake metadata, write a migration script for our product_sales table

AI Assistant: Based on the Delta Lake documentation and your Snowflake schema, here's a migration approach: [generates specific migration code]

API Integration Research​

User: /web REST API best practices data analytics 2024

AI Assistant: [Returns current API best practices]

User: /fetch https://docs.github.com/en/rest/guides/best-practices-for-integrators

AI Assistant: [Loads GitHub API best practices]

User: Help me design an API endpoint for our analytics dashboard based on these best practices

AI Assistant: Combining the search results and GitHub's best practices, here's an optimal API design: [detailed API specification]

Content Caching and Reuse​

Referencing Previous Fetches​

  • AI Assistant remembers content from previous /fetch commands in the current session
  • Reference earlier content without re-fetching URLs
  • Compare multiple sources loaded in the same session

Building Knowledge Base​

Session Example:

User: /fetch https://spark.apache.org/docs/latest/sql-performance-tuning.html
User: /fetch https://docs.databricks.com/optimizations/index.html
User: /web Spark performance optimization techniques 2024

User: Now compare all three sources and recommend optimization strategies for our ETL pipeline

AI Assistant: Analyzing the Spark documentation, Databricks optimization guide, and current best practices I found, here's a comprehensive optimization strategy: [integrated recommendations from all sources]

Troubleshooting​

Common Issues​

Rate Limit Exceeded​

Symptoms: Command returns rate limit error message

Solutions:

  1. Wait for limit reset (hourly or daily limits)
  2. Check current usage with /help command
  3. Plan command usage more strategically
  4. Prioritize essential searches and fetches

URL Fetch Failures​

Symptoms: /fetch command cannot access specified URL

Solutions:

  1. Verify URL accessibility in a web browser
  2. Check for typos in the URL
  3. Ensure proper protocol (https:// or http://)
  4. Try alternative URLs for the same content

Search Results Not Relevant​

Symptoms: /web command returns irrelevant or outdated results

Solutions:

  1. Refine search terms with more specific keywords
  2. Add date qualifiers (2024, latest, recent)
  3. Include technical terms and product names
  4. Try different query approaches and synonyms

Command Not Recognized​

Symptoms: Slash command doesn't execute or shows error

Solutions:

  1. Check command syntax using /help
  2. Verify spacing and parameter format
  3. Ensure command starts with forward slash (/)
  4. Try /help to confirm feature availability

Performance Optimization​

Efficient Command Usage​

  • Combine related queries to minimize command usage
  • Use specific search terms for more targeted results
  • Cache important information locally for repeated reference
  • Plan research sessions within rate limit windows

Content Processing Tips​

  • Prefer smaller, focused pages over large documents
  • Use specific documentation sections rather than entire sites
  • Check content accessibility before fetching
  • Bookmark frequently referenced URLs for easy access

Best Practices​

Strategic Usage​

Research Planning​

  • Identify key information needs before starting slash command usage
  • Prioritize essential searches within rate limits
  • Combine multiple approaches (web search + content fetch)
  • Document important findings for future reference

Workflow Integration​

  • Use slash commands early in analysis phases for context building
  • Fetch documentation before starting complex development work
  • Research best practices before implementing new techniques
  • Validate approaches with current industry standards

Effective Communication​

Clear Command Syntax​

  • Follow exact syntax shown in examples
  • Include complete URLs for fetch commands
  • Use descriptive search terms for web queries
  • Check spelling and formatting before executing

Follow-up Questions​

  • Ask specific questions about fetched content
  • Request comparisons between multiple sources
  • Seek practical applications of research findings
  • Get implementation guidance based on research

Collaboration and Sharing​

Team Knowledge Sharing​

  • Document useful URLs and search queries for team use
  • Share research findings with team members
  • Create knowledge bases from frequently fetched content
  • Establish team standards for research workflows

Project Integration​

  • Include research sources in project documentation
  • Reference external resources in code comments
  • Maintain links to relevant documentation
  • Update references as projects evolve

Summary​

You have successfully learned to use AI Assistant slash commands! Key achievements:

✅ Command Mastery: Understanding of /fetch, /web, and /help commands ✅ Usage Management: Awareness of rate limits and efficient usage strategies ✅ Research Workflows: Integration of external content with AI-powered analysis ✅ Content Integration: Effective use of fetched content and search results ✅ Troubleshooting: Solutions for common issues and optimization techniques ✅ Best Practices: Strategic planning and effective command usage patterns

AI Assistant slash commands significantly enhance your analysis capabilities by providing real-time access to external information, current trends, and comprehensive documentation, all integrated seamlessly with your data analysis workflow.