selan.aiDocs · Statusline
Claude Code statusline usage on a company account
Your Claude Code usage is on the line Selan adds to the bottom of Claude Code: your
company's 5-hour and 7-day rate-limit headroom, and your own spend against your own budget.
It installs itself, with no file to write and no flag to pass. selan usage
prints the same numbers without launching Claude Code. To install and sign in, start at
the docs.
What does the Claude Code statusline show?
Who you are, what you are running, and how much room you have left:
| Segment | What it tells you |
|---|---|
| Selan | The company you are signed in to and spending against. |
| Opus 5 (1M context) | The model actually serving this session, after any /model change, not what you launched with. |
| ⎇ main* | Your branch. The * means uncommitted changes (the one to check before letting seland loose). |
| ctx 18% | How full this session's context window is, as Claude Code reports it. It is left out when Claude Code does not send it. |
| 5h 32% | How much of your company's 5-hour rate limit is used. |
| 7d 39% | The same over 7 days. The slower one to recover from. |
| day $0.00/$50 | Your own spend against your own budget, in whichever period is tightest right now, so it may say week or month instead. |
Why do the 5h and 7d meters show my company's usage, not mine?
Because you do not have a personal rate-limit window. The meters are your company's headroom, and Selan picks the least-loaded account for each request. When they run hot, expect prompts to slow down, and start a long job later.
The meters follow the model in front of you. A Claude model shows your company's Anthropic accounts; a model on another provider shows that provider's own two windows. A model whose provider reports no window, such as one billed per token through a key, gets no 5h or 7d meter at all. A missing window is never drawn as 0%, because 0% would read as plenty of room.
Fleet numbers can sit a couple of minutes behind the dashboard. Your own spend is at most
a minute old: the statusline reuses its last reading for 60 seconds, and
selan usage asks afresh each time.
What does the budget meter on the statusline show?
Your own spend against your own limit, in whichever period (a day, a week or a month) is closest to its limit by the fraction used. With no limit set for you, there is no budget meter.
Spend is attributed to you personally, so one person running hot does not eat into someone else's allowance. Reaching your limit stops only what is billed per token. Your company's subscription seats keep serving you.
What happens at the limit is in limits and spend, and what an owner sees of each person's spend in spend per developer.
What do the colours on the meters mean?
Each bar fills with what is used, so the bar and the number beside it are the same reading. A bar turns amber at 70% and orange at 85%. From 85%, a rate-limit meter also shows when its window resets, in your own time zone.
How do I check Claude Code usage without the statusline?
Run selan usage. It prints the same numbers without launching Claude Code:
selan usageClaude's 5-hour and 7-day windows come first, then any provider whose windows the Claude figure leaves out, then your budget.
Codex has no statusline, so selan usage is how a Codex user reads the
same numbers. Every command is in the reference.
How do I turn the Selan statusline off?
Set SELAN_NO_STATUSLINE to any value.
If you already have a statusline of your own, in your own or a repository's Claude Code
settings, Selan leaves it alone. Passing your own --settings to Claude Code
also means no Selan statusline, as what an owner sets explains.