cli-harness-statusline

README

CLI Harness Statusline

A minimal and functional statusline for Cursor (CLI) Agnet & Claude Code

Installation + usage

One-liner (auto-detects Go vs release)

curl -sL https://raw.githubusercontent.com/5c077m4n/cli-harness-statusline/refs/heads/master/install.sh | bash

[!IMPORTANT] Always inspect scripts before piping them into your shell. You can review install.sh first.

Configuration

The statusline is configured via ~/.config/cli-harness-statusline/config.json. All fields are optional — any omitted field falls back to its default.

{
  "segments": {
    "model": { "disable": false }, // Show current AI model name
    "folder": { "disable": false }, // Show working directory name
    "git": { "disable": false }, // Show git branch and dirty state
    "worktree": { "disable": false }, // Show git worktree name
    "pr": { "disable": false }, // Show PR number and review state
    "agent": { "disable": false }, // Show active agent name
    "session": { "disable": false }, // Show session name (truncated)
    "context": { "disable": false }, // Show context usage bar and percentage
    "cost": { "disable": false }, // Show total cost, duration, lines
    "token": { "disable": false }, // Show input/output token counts
    "cache": { "disable": false }, // Show cache status and hit ratio
    "vim": { "disable": false }, // Show vim INSERT/NORMAL mode
    "autorun": { "disable": false }, // Show "auto" when autorun enabled
    "max": { "disable": false }, // Show "max" when max mode enabled
    "fastMode": { "disable": false }, // Show "fast" when fast mode enabled
    "effort": { "disable": false }, // Show effort level
    "thinking": { "disable": false }, // Show "think" when thinking enabled
    "rateLimit": { "disable": false }, // Show rate limit usage percentages
    "exceeds": { "disable": false } // Show 200k+ context warning marker
  },
  "padding": { "top": 0, "bottom": 0 } // Number of blank lines above/below the statusline
}

Screenshot

Benchmarks (using hyperfine)

See the latest results in CI run under the Benchmarks step summary.