openclaw - 💡(How to fix) Fix Context usage % display bug: cumulative cacheRead/cacheWrite divided by context window size instead of current conversation tokens [2 comments, 2 participants]

TL;DR

Update the status card computation to reflect the current conversation's token count relative to the context window size.

Guidance

• Review the internal-Difqhfo-.js file to identify where the context usage is calculated and update the formula to use the current conversation's token count instead of cumulative cache reads.
• Consider adding a new variable to track the current conversation's token count and use this value in the context usage calculation.
• Verify that the contextTokens value is being used correctly and not being overwritten or modified elsewhere in the code.
• Test the updated calculation with various session scenarios to ensure accuracy.

Example

// Pseudo-code example, actual implementation may vary
const currentConversationTokenCount = getConversationTokenCount(); // implement this function to get the current conversation's token count
const contextWindowSize = 200000; // model's context window size
const contextUsage = (currentConversationTokenCount / contextWindowSize) * 100;

Notes

The provided information suggests that the issue is specific to the status card rendering in the internal-Difqhfo-.js file, and updating the calculation should resolve the issue. However, further testing and verification are necessary to ensure the fix works as expected.

Recommendation

Apply workaround: update the status card computation to reflect the current conversation's token count relative to the context window size, as this directly addresses the root cause of the issue.