[FEATURE] Status line token arrows are counterintuitive — suggest adding descriptive labels
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
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
Problem
The current status line displays token counts like this:
Scampering… (1m 14s · ↓ 1.8k tokens · thinking)
The ↓ arrow is used for input tokens, but this is counterintuitive. Most users associate arrows with network traffic conventions:
- ↑ = upload = client sending data to the server/model
- ↓ = download = server/model sending data to the client
Using ↓ for input tokens is the opposite of this expectation.
Suggestion
- Reverse the arrow directions so they match standard network conventions:
- ↑ for input tokens (client → model)
- ↓ for output tokens (model → client)
- Add short descriptive labels to make it immediately clear without needing to learn the convention:
Scampering… (1m 14s · in ↑ 1.8k · out ↓ 0.3k tokens · thinking)
Why
New users shouldn't need to look up what the arrows mean. The labels in and out make it self-explanatory at a glance, and
aligning with the upload/download convention reduces cognitive friction.
Proposed Solution
Suggestion
- Reverse the arrow directions so they match standard network conventions:
- ↑ for input tokens (client → model)
- ↓ for output tokens (model → client)
- Add short descriptive labels to make it immediately clear without needing to learn the convention:
Scampering… (1m 14s · in ↑ 1.8k · out ↓ 0.3k tokens · thinking)
Alternative Solutions
_No response_
Priority
Medium - Would be very helpful
Feature Category
CLI commands and flags
Use Case Example
_No response_
Additional Context
_No response_
Proposed Solution
Suggestion
- Reverse the arrow directions so they match standard network conventions:
- ↑ for input tokens (client → model)
- ↓ for output tokens (model → client)
- Add short descriptive labels to make it immediately clear without needing to learn the convention:
Scampering… (1m 14s · in ↑ 1.8k · out ↓ 0.3k tokens · thinking)
Alternative Solutions
_No response_
Priority
Medium - Would be very helpful
Feature Category
CLI commands and flags
Use Case Example
_No response_
Additional Context
Re-opened https://github.com/anthropics/claude-code/issues/26920
This issue has 2 comments on GitHub. Read the full discussion on GitHub ↗