Skip to content
Rival Docs

Analytics

Rival tracks every execution that runs against your tools and surfaces that data in the Dashboard . Analytics give you a clear picture of how your tools are being used, how much they cost, and where problems occur.

What analytics covers

The analytics section in Dashboard shows:

  • Execution history - a log of every invocation, in reverse chronological order
  • Credit consumption - how many credits each tool has consumed over time
  • Execution volume - how many times each tool has been called, with trends over time
  • Success and failure rates - the proportion of executions that completed successfully versus those that returned an error

These metrics are available at both the workspace level (across all your tools) and per individual tool.

What each execution log entry shows

Every execution is recorded as a log entry. Each entry includes:

  • The tool that was called and the version invoked
  • The input payload sent by the caller
  • The output returned by the handler
  • The number of credits consumed
  • Whether the execution succeeded or failed, and the status code

This level of detail makes logs genuinely useful for debugging. If a caller reports unexpected behavior, you can pull up the exact execution - including its input and output - and trace what happened.

Using analytics for debugging

When a tool fails, the execution log is the first place to look. The input and output recorded for a failed execution tell you exactly what came in and what your handler returned (or threw). You do not need to reproduce the issue locally - the data is already captured.

For recurring failures, the failure rate trend makes it easy to spot whether a problem started with a specific version or correlates with particular input patterns.

Understanding credit consumption

Credits are deducted automatically based on compute time. The analytics dashboard shows credit consumption broken down by tool, which makes it straightforward to see which tools are the most resource-intensive and to optimize accordingly.

If a tool’s credit consumption is higher than expected, the execution logs - which show per-execution credit cost alongside the input and output - can help you identify which calls are expensive and why.