ccstatusline Total counter is misleading: inflates by re-adding cached tokens every request

Preflight Checklist

• [x] I have searched existing requests and this feature hasn't been requested yet
• [x] This is a single feature request (not multiple features)

Problem Statement

Summary

The Total token counter in the ccstatusline status line accumulates cached tokens
on every API request, making it appear as if massive token consumption is occurring
when it is not.

Example

After a 20-minute conversation:

• Total: 2.0M
• Cached: 2.0M
• In: 39.7k (actual new content)
• Cost: $0.60 (actual spend)

The same ~25k system prompt is re-sent from cache on every request. Each re-use
adds to the Total counter, so it balloons to millions while actual cost barely moves.

Problem

Total should reflect unique or meaningful token usage, not the cumulative sum of
cache reads across every request. For a new user, seeing 2.0M total is alarming
and gives a completely false picture of resource consumption.

Proposed Solution

Suggestion

• Exclude cached tokens from Total, or
• Add a separate Unique or New counter that excludes cache reads, or
• At minimum, document that Total includes repeated cache reads.

Alternative Solutions

_No response_

Priority

Low - Nice to have

Feature Category

API and model interactions

Use Case Example

Model: Sonnet 4.6 Total: 2.5M Out: 29 In: 45.8k Cached: 2.4M
Weekly: 2.0% Session: 3.0% Cost: $5.75 Weekly Reset: 5d 8hr 51m Block: 1hr 58m
Ctx: 45.8k Ctx Used: 23.0% Mem: 7.2G/11.7G Thinking: low Reset: 3hr 1m
This is ridiculous for a 45 min sparse session.

Additional Context

_No response_