Skip to content
Back to skills

Statusline

ASecurity

Install the skrrt Claude Code status line, a Catppuccin Mocha pill bar showing model, folder, git branch, lines changed, context use, and the 5-hour and weekly rate limits. Use when the user asks to install, set up, update, or restore the skrrt status line.

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 22, 2026
ai-agentspythonbashrailsgit

Works with

  • claude code
  • terminal

Security analysis

A100/100

Pro scans all 6 files and shows the line behind each finding

Scanned October 22, 2026

npx -y skills add skrrt-sh/skills --skill statusline --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Statusline?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Statusline
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/skrrt-sh-statusline/badge)](https://www.skillsdirectory.com/skills/skrrt-sh-statusline)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: statusline
description: Install the skrrt Claude Code status line, a Catppuccin Mocha pill bar showing model, folder, git branch, lines changed, context use, and the 5-hour and weekly rate limits. Use when the user asks to install, set up, update, or restore the skrrt status line.
allowed-tools: Bash(${CLAUDE_SKILL_DIR}/scripts/install-statusline.sh:*)
---

# Install the Status Line

## Process

1. Ask the user for a context budget, the token count the context pill treats as full, unless they
   already named one. Make clear it is visual feedback on how efficiently the session uses its
   context: it never caps the model's real window, and nothing breaks past it. Offer:
   - **100k**: strict. Quality drops are reported from about 100k even on 1M-window models.
   - **200k**: balanced. Where Claude Code compacted before 1M windows.
   - **256k** (default): the honest ceiling for decent results. Heavy users of 1M-window models
     report drops between 250k and 400k. Yellow starts at 128k, where quality drops are commonly
     reported.
   - A custom token count.
2. Run `${CLAUDE_SKILL_DIR}/scripts/install-statusline.sh <tokens>`, for example `200000`. It copies
   [assets/statusline.py](assets/statusline.py) to `statusline/statusline.py` in the Claude config
   directory and sets `statusLine` in that directory's `settings.json`, passing the budget to the
   script. To change only the budget later, run it again with the new number.
3. Report `INSTALLED`, `SETTINGS`, and `BUDGET`. If it printed `BACKUP`, name the backup of the
   replaced script. If it printed `PREVIOUS_STATUSLINE`, show the replaced setting so the user can
   restore it.
4. Tell the user the bar appears on the next refresh and its icons need a Nerd Font in the
   terminal.
5. Walk the user through the bar, left to right, using [Reading the bar](#reading-the-bar).

## Reading the bar

- Model, then folder.
- Git branch. Its `+N -M` tail is the uncommitted diff against the last commit, untracked files
  included: what the next commit would ship, whoever wrote it.
- Context: tokens in use against the chosen soft budget. Green below half, yellow from half. Past
  the budget it turns red and reads `new session`: the session still works, but it is time to start
  a fresh one. Outside a git repo it carries the `+N -M` tail instead: lines Claude changed in this
  session, so it lives and resets with the context.
- `5h` and `wk`: rate limit used, with time until reset. Red from 80%.

## Guardrails

- Install only when asked. To change colors or segments, edit the installed script, not the asset,
  unless the user is changing this skill.
- On a missing `python3` or `jq`, report the error and stop; do not edit `settings.json` by hand.

Files in this skill

  • SKILL.md2.7 KB
  • agents/openai.yaml218 B
  • assets/statusline.py5.1 KB
  • evals/evals.json2.4 KB
  • evals/trigger-evals.json1.7 KB
  • scripts/install-statusline.sh1.5 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.