claude-vibeline 1.1.0

Claude Code statusline with real subscription usage data

Claude Vibeline

A custom statusline for Claude Code that shows session details, prompt cache status, and real subscription usage data — designed for Pro, Max, and Team users.

Unlike token-based cost trackers, Claude Vibeline shows your actual rate limit utilization and reset countdowns as reported by Anthropic, alongside context window usage, effort level, and a live cache countdown.

my-project │ Opus 4.6 (high) │ cache ✓ 4m │ ctx 200k [###-----] 42% │
sess [##------] 19% 3h12m │ week [--------] 3% 5d20h │ extra 1.23/20€ 7d0h

Sections wrap to multiple lines based on --columns width, with a trailing │ to indicate continuation.

Note: This project was developed almost entirely through AI-assisted coding with Claude Code, with human oversight over all design decisions, architecture, and code review.

Features

From Claude Code's session data:

• Project & model - project name, active model, and effort level (resolved from session transcript with settings.json fallback)
• Context window - how much of the context window is used, with size indicator (e.g. 200k, 1M)
• Prompt cache - tracks Anthropic's 5-minute prompt cache TTL for subscription users. Computed from user message timestamps in the session transcript, with a background countdown that refreshes without waiting for the next invocation:
  • ✓ 4m — cache is warm, 4 minutes remaining
  • ⚠ 47s — cache is warm but below the low-threshold warning
  • ✗ — cache has expired
  • ! prefix — cache expired between user messages earlier in the session (cache gap detected)

From Anthropic's OAuth API (subscription rate limits, not per-token costs):

• Session limit - 5-hour rate limit utilization with reset countdown
• Weekly limit - 7-day rate limit utilization with reset countdown
• Per-model limits - weekly Opus and Sonnet limits with reset countdowns, shown when applicable
• Extra usage - spend against your monthly extra usage cap with reset countdown, shown only if enabled

Every section is individually toggleable via CLI flags.

Installation

Available on PyPI. We recommend using uv.

To install:

uv tool install claude-vibeline

To update:

uv tool upgrade claude-vibeline

Setup

Add to ~/.claude/settings.json:

{
  // ...
  "statusLine": {
    "type": "command",
    "command": "claude-vibeline"
  }
}

Options

Flag	Description
--columns N	Terminal width in characters (default: 80)
--bar-width N	Progress bar width in characters (default: 8)
--currency S	Currency symbol for extra usage (default: €)
--no-project	Hide project name
--no-model	Hide model and effort level
--no-cache	Hide prompt cache status
--no-refresh	Disable background cache timer refresh
--no-context	Hide context window usage
--no-usage	Skip fetching usage data entirely
--no-session	Hide session (5h) usage
--no-weekly	Hide weekly (7d) usage
--no-opus	Hide weekly Opus usage
--no-sonnet	Hide weekly Sonnet usage
--no-extra	Hide extra usage spend
--debug	Log each statusline output to debug file

Example with customizations:

{
  // ...
  "statusLine": {
    "type": "command",
    "command": "claude-vibeline --bar-width 5 --currency $ --no-cache --no-opus --no-sonnet"
  }
}

my-project │ Opus 4.6 (high) │ ctx 200k [##---] 42% │ sess [#----] 19% 3h12m │
week [-----] 3% 5d20h │ extra 1.23/20$ 7d0h

If everything fits on a single line, no trailing │ is shown.

Usage data

Usage data is fetched from an Anthropic OAuth endpoint (see Limitations). It requires a valid OAuth token from a Claude Pro, Max, or Team subscription.

Responses are cached locally for 60 seconds. Cached usage data is used when the token expires or the API is unavailable. If no token or cache exists, usage sections are omitted. Use --no-usage to disable API calls entirely.

Limitations

• Undocumented APIs — the OAuth usage endpoint and the cache timer refresh mechanism are both undocumented and may break without notice.
• Limited stdin data — the statusline process receives only a JSON blob on stdin. Claude Code's own CLI arguments (e.g. --model) and internal environment variables are not accessible.
• Effort level is inferred — effort is not provided in stdin. It is resolved from the session transcript by scanning for /model and /effort commands, with a settings.json fallback shown with ? suffix. After session resume, effort resets to the ? fallback until /effort or /model is used.
• No session fork support — transcript parsing does not distinguish between session forks. Multiple instances sharing a transcript will see each other's messages, and their session caches will race.

Development

Requires uv.

Clone and setup:

git clone https://github.com/hstojanovic/claude-vibeline.git
cd claude-vibeline
uv sync

Run checks:

uv run ruff format --check
uv run ruff check
uv run ty check
uv run pytest --cov

Build:

uv build

License

This project is licensed under the MIT license.

---

Claude Vibeline is an independent project and is not affiliated with or endorsed by Anthropic.