statusline

name	statusline
description	Enable, configure, or disable the Magician CLI status line ("Magician Claude CLI UI") — a lightweight, always-on bar showing live context %, a context-rot warning, a token-flow sparkline, model · git · cost, the active skill/workflow/loop, the reasoning effort/mode, the bundled-lore state, and the output-brevity "voice" level. Also sets the output-brevity voice (warrior/scribe/bard). Use when the user says "enable/turn on the magician UI / status line / status bar", "show my context/tokens/effort in the console", "configure/change what the bar shows", "what should the status line display", "turn off / disable the status line", or "make output shorter/leaner / set the voice / reduce token cost".
allowed-tools	Bash(magician-ui:*), AskUserQuestion, Read
argument-hint	["enable · disable · status · set <components>"]

/statusline — Magician CLI status line

A native Claude Code status line rendered by magician. It runs locally, consumes zero API tokens, updates on each message (debounced), and helps you catch context rot before it bites. It's driven by the bundled magician-ui CLI, which edits ~/.claude/settings.json safely (timestamped backup → validate → atomic write; it never leaves settings broken and only ever touches the statusLine key).

Components (user-configurable)

key	shows
context	color-coded usage bar + used% + used/size tokens (green <70 · yellow 70–89 · red ≥90)
rot	a ⚠ at ≥80% and 🔴 at ≥92% so you notice before compaction
spark	a ▁▂▃▅▇ sparkline of recent context% (the token-flow stream)
meta	model · git branch · session cost
skill	the active magician skill / workflow / running-agent count / loop round
effort	🧠 the live reasoning effort (low/medium/high/xhigh/max) — the default shows on open and tracks /effort changes automatically (from Claude Code's effort.level) — or the magician mode you set, e.g. ultracode (which otherwise reports as xhigh); say "set mode to ultracode" / "exit ultracode" to change it
lore	📚 whether magician's bundled stack lore is shaping the session — lore:N (N cores injected) or lore:off
voice	🗣 the output-brevity voice — voice:warrior (leanest) · voice:scribe (default) · voice:bard (standard); set with magician-ui voice <level>

Actions

• Enable — first confirm which components the user wants (default: all). Use AskUserQuestion (multi-select) unless they already said (e.g. "just rot and context"). Then:

  magician-ui enable --all              # everything
  magician-ui enable --only context,rot # a chosen subset
  

• Change what shows (no re-enable needed):

  magician-ui set context,rot,spark
  

• Status — magician-ui status (state, components, whether it's wired into settings).
• Disable — magician-ui disable (removes only magician's statusLine; records the opt-out so it isn't suggested again; re-enable anytime on request).
• Auto mode (real autonomy) — magician-ui automode turns on Claude Code's auto permission mode (defaultMode: auto + CLAUDE_CODE_ENABLE_AUTO_MODE=1, required on Vertex/Bedrock/Foundry). Its classifier auto-approves reads + request-aligned work and gates writes/deploys/force-push/destructive ops — the true "reads proceed, writes gate." A plugin can't switch the mode of a live session, so restart to enter it; automode --off reverts. Falls back to Manual if the account/model doesn't support auto.
• Read-only auto-approve (acceptEdits fallback) — magician-ui allow merges a read-only allow-list (Read/Grep/Glob/LS + read-only git + kg/ctx + jira/confluence reads + test/lint/build runners + gh reads) into settings.json so non-auto sessions don't prompt per read; jira/confluence writes, commit/push/PR/delete still gate. Applied on install/upgrade; magician-ui allow --off. (Jira/Confluence use magician's MCP-free CLIs — magician nudges sessions off any ambient Atlassian MCP.) See lore/autonomy.md.
• Voice — output brevity (lower token cost, no quality loss) — magician-ui voice warrior|scribe|bard sets how wordy responses are. Output tokens cost several× input, so leaner output is the cheapest saving. Levels least→most wordy: warrior (minimal but complete), scribe (the default — leaner than usual), bard (standard/native). SessionStart injects a brevity directive for warrior/scribe that cuts filler (preambles, recaps, restating the request) while keeping all substance and code/commands/errors verbatim — it never compresses prose into fragments or jargon. magician-ui voice status shows the current level. Overrides (first match wins): env MAGICIAN_VOICE → per-project .magician/voice → this setting → default scribe. Takes effect next session start; the 🗣 voice: chip shows it live.

Rules

• The bar is suggested once on session start when never configured; after that magician stays quiet and only acts when the user asks (reconcile records the state). Respect a decline — don't re-nudge.
• Enabling/disabling changes ~/.claude/settings.json; it hot-reloads, so the bar appears/updates within a message or two — no restart needed.
• Keep it lightweight: recommend a smaller component set if the user wants a minimal bar (e.g. context,rot).
• Never hand-edit settings.json for this — always go through magician-ui, which backs up + validates.

Completion Signal

"Magician CLI UI <enabled (components: …) | updated | disabled>. It runs locally (no tokens) and hot-reloads."