Log History

Monitor your Fetch Hive interactions with log history

Viewing activity log history

Overview

Log history in Fetch Hive provides comprehensive insights into all interactions across your workspace. Whether you're debugging issues, optimizing performance, or analyzing usage patterns, the log history gives you detailed information about every operation performed within the platform.

Available Log Types

You can access logs for the following resources:

  • Prompts: View all executions of your AI prompts

  • Endpoints: Track usage of both Prompt and Workflow endpoints

  • Workflows: View all executions of your Workflow steps

  • Datasets: Monitor dataset operations and updates

  • Agents: Review agent interactions and responses

  • Fine Tuning: Track fine-tuning job progress and results

Log Details

Each log entry provides extensive information about the interaction, including:

Common Fields

  • Output: The result or response from the interaction

  • Inputs: Parameters and data provided to trigger the interaction

  • Credit Usage: Amount of credits consumed

  • Duration: Time taken to complete the operation

  • Provider: The service provider used (e.g., Exa, Google Search, OpenAI, Claude)

  • Owner: User who initiated the interaction

  • Status: Current state of the operation (e.g., Completed, Error, Running)

  • Timestamp: When the interaction occurred

AI-Specific Fields

For AI Prompt Steps or Prompts, additional information includes:

  • Model: The specific AI model used

  • Tokens: Number of tokens consumed

  • LLM Charge: Cost associated with the language model usage

Step-Specific Data

Workflow steps may include additional custom data fields depending on the step type, providing deeper insights into the operation.

Using Log History

Debugging and Troubleshooting

  • Identify failed operations and their error messages

  • Review input parameters that led to specific outcomes

  • Track the sequence of events in complex workflows

Performance Optimization

  • Monitor response times and resource usage

  • Analyze token consumption patterns

  • Identify opportunities for cost optimization

Testing and Iteration

Log history is invaluable for:

  • Comparing different versions of prompts or workflows

  • Understanding which approaches yield better results

  • Fine-tuning parameters based on historical performance

  • Validating changes before deploying to production

Usage Analytics

  • Track usage patterns across different resources

  • Monitor credit consumption by resource type

  • Analyze user engagement and adoption

Viewing Logs

  1. Navigate to the respective resource section (Prompts, Endpoints, Datasets, etc.)

  2. Click Activity Logs in the top right of the navbar

Accessing activity logs for Workflows

Best Practices

  1. Regular Monitoring: Review logs periodically to catch issues early

  2. Performance Tracking: Use log data to optimize resource usage

  3. Documentation: Reference logs when documenting successful patterns

  4. Iteration: Use historical data to inform improvements

  5. Cost Management: Monitor credit usage to optimize expenses

Last updated

Was this helpful?