Claude Code Status Line - Complete Guide: all fields, config, ready-to-use scripts

A developer published a guide to Claude Code's status line feature, a customizable bar at the bottom of the terminal that displays real-time session data such as model name, context usage, cost, and code velocity. The guide includes ready-to-use scripts and configuration examples, emphasizing that the feature runs entirely locally with no API usage.

Claude Code Status Line - Complete Guide: all fields, config, ready-to-use scripts A persistent, customizable bar at the bottom of Claude Code that shows real-time session data. What Is It? https://gist.github.com/starred.atom what-is-it Quick Start https://gist.github.com/starred.atom quick-start Want Mine? https://gist.github.com/starred.atom want-mine Available Data Fields https://gist.github.com/starred.atom available-data-fields Your Script Superpowers https://gist.github.com/starred.atom your-script-superpowers Ready-to-Use Scripts https://gist.github.com/starred.atom ready-to-use-scripts Full JSON Schema https://gist.github.com/starred.atom full-json-schema Tips and Tricks https://gist.github.com/starred.atom tips-and-tricks All Claude Code Settings https://gist.github.com/starred.atom all-claude-code-settings The status line is an info bar pinned to the bottom of your Claude Code window . Think of it like the status bar in VS Code or the bottom of a Google Doc — it just sits there showing you useful stuff while you work. What can it show? - Which AI model you're using - Your project name and git branch - How much of your conversation memory is used up context % - How much the session has cost so far - How long you've been working - Lines of code added / removed - …and more How does it work? You tell Claude what you want to see, and it builds the status bar for you. No coding required on your part — just describe it in plain English. Does it cost anything? Nope. It runs entirely on your machine. Zero API usage. Just type a /statusline command describing what you want: /statusline show model name and context percentage with a progress bar Claude generates the script, saves it, and wires up settings for you. Done. More examples: /statusline show repo name, git branch, context bar, and model /statusline show cost and session duration with model name /statusline show git branch with colored staged/modified file counts /statusline two lines: model and branch on top, color-coded context bar with cost on bottom To remove it: /statusline clear Here's the exact status line I use daily. Fully color-coded with icons and a gradient context bar. What it looks like: | Element | Example | Color | |---|---|---| | Repo name | EasyClaw | Bold yellow | | Branch | 🌿 main | Bold cyan | | Context bar | ⚡ █████████░░░░░░░░░░░ 47% | True RGB gradient green → yellow → red + dynamic emoji | | Context emoji | 🟢 → ⚡ → 🔥 → 🚨 | Changes at 20% / 70% / 90% | | Cost | $0.47 | Yellow | | Code velocity | +156 -23 | Green adds, red deletes | | Model | 🤖 Opus 4.6 | Magenta | | Separators | | | Dim gray | One command to get it: /statusline single line, fully color-coded with truecolor RGB gradient: repo name bold yellow, 🌿 leaf icon + git branch in bold cyan with parentheses, 20-block context bar using 24-bit RGB gradient from green 0,200,80 through yellow 220,200,0 to red 220,40,20 with dark gray 60,60,60 empty blocks, dynamic emoji that changes with usage 🟢 under 20%, ⚡ 20-69%, 🔥 70-89%, 🚨 90%+ , percentage colored by usage level, session cost in yellow, code velocity showing +lines in green and -lines in red, 🤖 robot icon + model name in magenta, dim gray pipe separators between all elements Or copy the script directly: bash /usr/bin/env bash Claude Code status line: RGB gradient, dynamic emoji, cost, code velocity input=$ cat ── Colors ── CYAN='\033 36m' GREEN='\033 32m' YELLOW='\033 33m' RED='\033 31m' MAGENTA='\033 35m' DIM='\033 2m' BOLD='\033 1m' RESET='\033 0m' ── Truecolor helper ── rgb { printf '\033 38;2;%d;%d;%dm' "$1" "$2" "$3"; } ── Parse JSON fields ── model=$ echo "$input" | jq -r '.model.display name // "Unknown"' used=$ echo "$input" | jq -r '.context window.used percentage // empty' cost=$ echo "$input" | jq -r '.cost.total cost usd // 0' lines add=$ echo "$input" | jq -r '.cost.total lines added // 0' lines del=$ echo "$input" | jq -r '.cost.total lines removed // 0' cwd=$ echo "$input" | jq -r '.workspace.current dir // .cwd // ""' ── Git info ── branch="" repo="" if -n "$cwd" ; then branch=$ git -C "$cwd" --no-optional-locks symbolic-ref --short HEAD 2 /dev/null repo=$ basename "$ git -C "$cwd" --no-optional-locks rev-parse --show-toplevel 2 /dev/null " 2 /dev/null fi ── Context bar: RGB gradient, full blocks only ── BAR WIDTH=20 if -n "$used" ; then used int=$ printf '%.0f' "$used" Round to nearest block filled=$ used int BAR WIDTH + 50 / 100 bar="" for i=0; i<BAR WIDTH; i++ ; do pos=$ i 100 / BAR WIDTH - 1 if "$pos" -le 50 ; then r=$ 0 + 220 pos / 50 g=200 b=$ 80 - 80 pos / 50 else adj=$ pos - 50 r=220 g=$ 200 - 160 adj / 50 b=$ 0 + 20 adj / 50 fi if "$i" -lt "$filled" ; then bar="${bar}$ rgb $r $g $b █" else bar="${bar}\033 38;2;60;60;60m░" fi done bar="${bar}${RESET}" if "$used int" -ge 90 ; then status emoji="🚨" elif "$used int" -ge 70 ; then status emoji="🔥" elif "$used int" -ge 20 ; then status emoji="⚡" else status emoji="🟢"; fi if "$used int" -ge 90 ; then pct color="$RED" elif "$used int" -ge 70 ; then pct color="$YELLOW" else pct color="$GREEN"; fi ctx part="${status emoji} ${bar} ${pct color}${used int}%${RESET}" else ctx part="🟢 \033 38;2;60;60;60m░░░░░░░░░░░░░░░░░░░░${RESET} --%" fi ── Cost ── cost part="${YELLOW}$ printf '$%.2f' "$cost" ${RESET}" ── Code velocity ── velocity="${GREEN}+${lines add}${RESET} ${RED}-${lines del}${RESET}" ── Single line ── out="" -n "$repo" && out="${BOLD}${YELLOW}${repo}${RESET}" -n "$branch" && out="${out:+$out }${BOLD}${CYAN}🌿 ${branch} ${RESET}" out="${out:+$out ${DIM}|${RESET} }${ctx part}" out="${out} ${DIM}|${RESET} ${cost part}" out="${out} ${DIM}|${RESET} ${velocity}" out="${out} ${DIM}|${RESET} ${MAGENTA}🤖 ${model}${RESET}" printf '%b' "$out" Every time your script runs, it receives JSON via stdin . Here is everything you can display. | Field | What It Shows | Example Value | |---|---|---| model.id | Model identifier | "claude-opus-4-6" | model.display name | Friendly model name | "Opus" | session id | Unique session ID | "abc123..." | session name | Custom name via /rename | "my-feature-work" | version | Claude Code version | "2.1.90" | | Field | What It Shows | Example Value | |---|---|---| cwd | Current working directory | "/Users/me/project" | workspace.current dir | Same as cwd preferred | "/Users/me/project" | workspace.project dir | Where Claude Code was launched | "/Users/me/project" | workspace.added dirs | Extra dirs added via /add-dir | | workspace.git worktree | Git worktree name if in one | "feature-xyz" | | Field | What It Shows | Example | |---|---|---| context window.used percentage | % of context used | 30 | context window.remaining percentage | % of context remaining | 70 | context window.context window size | Max tokens 200K or 1M | 200000 | context window.total input tokens | Cumulative input tokens | 15234 | context window.total output tokens | Cumulative output tokens | 4521 | context window.current usage.input tokens | Input tokens in last API call | 8500 | context window.current usage.output tokens | Output tokens from last call | 1200 | context window.current usage.cache creation input tokens | Tokens written to cache | 5000 | context window.current usage.cache read input tokens | Tokens read from cache | 2000 | exceeds 200k tokens | Last response exceeded 200K | false | | Field | What It Shows | Example Value | |---|---|---| cost.total cost usd | Total session cost USD | 0.01234 | cost.total duration ms | Wall-clock time since start | 45000 | cost.total api duration ms | Time waiting for API responses | 2300 | cost.total lines added | Lines of code added | 156 | cost.total lines removed | Lines of code removed | 23 | | Field | What It Shows | Example Value | |---|---|---| rate limits.five hour.used percentage | 5-hour usage % | 23.5 | rate limits.five hour.resets at | Reset time Unix epoch | 1738425600 | rate limits.seven day.used percentage | 7-day usage % | 41.2 | rate limits.seven day.resets at | Reset time Unix epoch | 1738857600 | | Field | When It Appears | Example Value | |---|---|---| vim.mode | When vim mode is enabled | "NORMAL" or "INSERT" | output style.name | Current output style | "default" | agent.name | When using --agent flag | "security-reviewer" | worktree.name | During --worktree sessions | "my-feature" | worktree.path | Worktree directory path | "/path/to/worktree" | worktree.branch | Worktree git branch | "worktree-my-feature" | worktree.original cwd | Dir before entering worktree | "/path/to/project" | worktree.original branch | Branch before entering worktree | "main" | transcript path | Path to conversation transcript | "/path/to/transcript" | Each echo / print creates a separate row: echo "Line 1: Model and branch info" echo "Line 2: Context bar and cost" +-----------------------------------------------------------------+ | Opus EasyClaw | main | | ...... 40% | $0.23 | 12m 5s | +-----------------------------------------------------------------+ Use escape codes for colored output: GREEN='\033 32m' YELLOW='\033 33m' RED='\033 31m' CYAN='\033 36m' RESET='\033 0m' echo -e "${GREEN}All good${RESET} | ${RED}Warning${RESET}" Use OSC 8 escape sequences requires iTerm2, Kitty, or WezTerm : printf '%b' "\e 8;;https://github.com/user/repo\aRepo Name\e 8;;\a" Hold Cmd macOS or Ctrl Windows/Linux and click to open. Copy any of these and pass them to /statusline , or just describe what you want in plain English. Opus ....... 30% bash /bin/bash input=$ cat MODEL=$ echo "$input" | jq -r '.model.display name' PCT=$ echo "$input" | jq -r '.context window.used percentage // 0' | cut -d. -f1 BAR WIDTH=10 FILLED=$ PCT BAR WIDTH / 100 EMPTY=$ BAR WIDTH - FILLED BAR="" "$FILLED" -gt 0 && printf -v FILL "%${FILLED}s" && BAR="${FILL// /▓}" "$EMPTY" -gt 0 && printf -v PAD "%${EMPTY}s" && BAR="${BAR}${PAD// /░}" echo " $MODEL $BAR $PCT%" EasyClaw main ███░░░░░░░ 30% Claude Opus 4 bash /bin/bash input=$ cat model=$ echo "$input" | jq -r '.model.display name // "Unknown Model"' used=$ echo "$input" | jq -r '.context window.used percentage // empty' if -n "$used" ; then used int=$ printf '%.0f' "$used" filled=$ used int / 10 empty=$ 10 - filled bar="" for i in $ seq 1 $filled ; do bar="${bar}█"; done for i in $ seq 1 $empty ; do bar="${bar}░"; done ctx part=" ${bar} ${used int}%" else ctx part=" ░░░░░░░░░░ --%" fi cwd=$ echo "$input" | jq -r '.workspace.current dir // .cwd // ""' branch="" repo="" if -n "$cwd" ; then branch=$ git -C "$cwd" --no-optional-locks symbolic-ref --short HEAD 2 /dev/null repo=$ basename "$ git -C "$cwd" --no-optional-locks rev-parse --show-toplevel 2 /dev/null " 2 /dev/null fi out="" -n "$repo" && out="$repo" -n "$branch" && out="${out:+$out } ${branch} " out="${out:+$out }${ctx part} ${model}" printf '%s' "$out" Opus EasyClaw | main +2 ~5 bash /bin/bash input=$ cat MODEL=$ echo "$input" | jq -r '.model.display name' DIR=$ echo "$input" | jq -r '.workspace.current dir' GREEN='\033 32m'; YELLOW='\033 33m'; RESET='\033 0m' if git rev-parse --git-dir /dev/null 2 &1; then BRANCH=$ git branch --show-current 2 /dev/null STAGED=$ git diff --cached --numstat 2 /dev/null | wc -l | tr -d ' ' MODIFIED=$ git diff --numstat 2 /dev/null | wc -l | tr -d ' ' GIT STATUS="" "$STAGED" -gt 0 && GIT STATUS="${GREEN}+${STAGED}${RESET}" "$MODIFIED" -gt 0 && GIT STATUS="${GIT STATUS}${YELLOW}~${MODIFIED}${RESET}" echo -e " $MODEL ${DIR /} | $BRANCH $GIT STATUS" else echo " $MODEL ${DIR /}" fi Opus $0.23 | 12m 5s bash /bin/bash input=$ cat MODEL=$ echo "$input" | jq -r '.model.display name' COST=$ echo "$input" | jq -r '.cost.total cost usd // 0' DURATION MS=$ echo "$input" | jq -r '.cost.total duration ms // 0' COST FMT=$ printf '$%.2f' "$COST" DURATION SEC=$ DURATION MS / 1000 MINS=$ DURATION SEC / 60 SECS=$ DURATION SEC % 60 echo " $MODEL $COST FMT | ${MINS}m ${SECS}s" Opus EasyClaw | main ...... 40% | $0.23 | 12m 5s Context bar changes color: green < 70% | yellow 70-89% | red 90%+ bash /bin/bash input=$ cat MODEL=$ echo "$input" | jq -r '.model.display name' DIR=$ echo "$input" | jq -r '.workspace.current dir' COST=$ echo "$input" | jq -r '.cost.total cost usd // 0' PCT=$ echo "$input" | jq -r '.context window.used percentage // 0' | cut -d. -f1 DURATION MS=$ echo "$input" | jq -r '.cost.total duration ms // 0' CYAN='\033 36m'; GREEN='\033 32m'; YELLOW='\033 33m'; RED='\033 31m'; RESET='\033 0m' if "$PCT" -ge 90 ; then BAR COLOR="$RED" elif "$PCT" -ge 70 ; then BAR COLOR="$YELLOW" else BAR COLOR="$GREEN"; fi FILLED=$ PCT / 10 ; EMPTY=$ 10 - FILLED printf -v FILL "%${FILLED}s"; printf -v PAD "%${EMPTY}s" BAR="${FILL// /█}${PAD// /░}" MINS=$ DURATION MS / 60000 ; SECS=$ DURATION MS % 60000 / 1000 BRANCH="" git rev-parse --git-dir /dev/null 2 &1 && BRANCH=" | $ git branch --show-current 2 /dev/null " echo -e "${CYAN} $MODEL ${RESET} ${DIR /}$BRANCH" COST FMT=$ printf '$%.2f' "$COST" echo -e "${BAR COLOR}${BAR}${RESET} ${PCT}% | ${YELLOW}${COST FMT}${RESET} | ${MINS}m ${SECS}s" Opus 5h: 23% | 7d: 41% bash /bin/bash input=$ cat MODEL=$ echo "$input" | jq -r '.model.display name' FIVE HR=$ echo "$input" | jq -r '.rate limits.five hour.used percentage // empty' SEVEN DAY=$ echo "$input" | jq -r '.rate limits.seven day.used percentage // empty' out=" $MODEL " -n "$FIVE HR" && out="$out 5h: $ printf '%.0f' "$FIVE HR" %" -n "$SEVEN DAY" && out="$out | 7d: $ printf '%.0f' "$SEVEN DAY" %" echo "$out" This is the complete JSON your script receives on stdin : { "cwd": "/current/working/directory", "session id": "abc123...", "session name": "my-session", "transcript path": "/path/to/transcript.jsonl", "model": { "id": "claude-opus-4-6", "display name": "Opus" }, "workspace": { "current dir": "/current/working/directory", "project dir": "/original/project/directory", "added dirs": , "git worktree": "feature-xyz" }, "version": "2.1.90", "output style": { "name": "default" }, "cost": { "total cost usd": 0.01234, "total duration ms": 45000, "total api duration ms": 2300, "total lines added": 156, "total lines removed": 23 }, "context window": { "total input tokens": 15234, "total output tokens": 4521, "context window size": 200000, "used percentage": 8, "remaining percentage": 92, "current usage": { "input tokens": 8500, "output tokens": 1200, "cache creation input tokens": 5000, "cache read input tokens": 2000 } }, "exceeds 200k tokens": false, "rate limits": { "five hour": { "used percentage": 23.5, "resets at": 1738425600 }, "seven day": { "used percentage": 41.2, "resets at": 1738857600 } }, "vim": { "mode": "NORMAL" }, "agent": { "name": "security-reviewer" }, "worktree": { "name": "my-feature", "path": "/path/to/.claude/worktrees/my-feature", "branch": "worktree-my-feature", "original cwd": "/path/to/project", "original branch": "main" } } Fields that may be absent: session name -- only when set via --name or /rename workspace.git worktree -- only inside a linked git worktree vim -- only when vim mode is enabled agent -- only when using --agent worktree -- only during --worktree sessions rate limits -- only for Claude.ai Pro/Max after first API response Fields that may be null: context window.current usage -- null before first API call context window.used percentage -- may be null early in session | Tip | Details | |---|---| When does it update? | After each assistant message, permission mode change, or vim mode toggle. Debounced at 300ms. | No API tokens used | Status line runs locally -- completely free. | Hides during prompts | Temporarily hidden during autocomplete, help menu, and permission prompts. | Can use any language | Bash, Python, Node.js -- anything that reads stdin and prints to stdout. | Test your script | Pipe sample JSON: echo '{"model":{"display name":"Opus"}}' | Every setting lives in settings.json . There are three levels, loaded in order later overrides earlier : | Level | Path | Scope | |---|---|---| | User | ~/.claude/settings.json | All projects on your machine | | Project | .claude/settings.json in repo root | Everyone on this repo | | Managed | ~/.claude/settings.managed.json | Enterprise IT policy read-only | | Setting | Type | Default | What It Does | |---|---|---|---| model | string | "claude-sonnet-4-6" | Default model for all sessions | availableModels | string | all | Restrict which models users can pick | effortLevel | "auto" | "low" | "medium" | "high" | varies | Opus 4.6 reasoning depth — lower = faster & cheaper | fastMode | boolean | false | 2.5x faster output, higher per-token cost same model | fastModePerSessionOptIn | boolean | false | Require /fast each session instead of persisting | language | string | — | Preferred language for Claude's responses | outputStyle | string | "default" | Response style: default , Explanatory , Learning , or custom | alwaysThinkingEnabled | boolean | false | Extended thinking on by default for all sessions | | Setting | Type | Default | What It Does | |---|---|---|---| statusLine.type | "command" | — | Must be "command" to enable a custom status line | statusLine.command | string | — | Shell command or script path that reads JSON stdin, prints status to stdout | statusLine.padding | number | 0 | Extra horizontal spacing characters around the status line content | | Setting | Type | Default | What It Does | |---|---|---|---| permissions.defaultMode | string | "default" | Default permission mode: default , plan , bypassPermissions , etc. | permissions.allow | string | | Tools always allowed without asking | permissions.deny | string | | Tools always denied | allowManagedPermissionRulesOnly | boolean | false | Managed only Block user/project permission rules | | Setting | Type | Default | What It Does | |---|---|---|---| spinnerTipsEnabled | boolean | true | Show tips in the spinner while Claude works | spinnerTipsOverride | object | — | Custom tip messages for the spinner | spinnerVerbs | object | — | Custom verbs shown in spinner progress | terminalProgressBarEnabled | boolean | true | Progress bar in terminals that support it Windows Terminal, iTerm2 | showTurnDuration | boolean | true | Show "Cooked for 1m 6s" after responses | prefersReducedMotion | boolean | false | Disable UI animations spinners, shimmer, flash for accessibility | companyAnnouncements | string | — | Managed only Startup messages, one picked randomly | | Setting | Type | Default | What It Does | |---|---|---|---| enabledPlugins | object | {} | Map of "plugin@marketplace": true to enable plugins | extraKnownMarketplaces | object | {} | Additional plugin marketplace sources | pluginConfigs | object | {} | Per-plugin configuration MCP user configs, etc. | skippedPlugins | string | | Plugins you declined when prompted | skippedMarketplaces | string | | Marketplaces you declined when prompted | strictKnownMarketplaces | string | — | Managed only Allowlist of permitted marketplaces | blockedMarketplaces | string | — | Managed only Blocked marketplace sources | pluginTrustMessage | string | — | Managed only Custom message on plugin trust warning | | Setting | Type | Default | What It Does | |---|---|---|---| enableAllProjectMcpServers | boolean | false | Auto-approve all MCP servers in the project | enabledMcpjsonServers | string | | Approved MCP servers from .mcp.json | disabledMcpjsonServers | string | | Rejected MCP servers from .mcp.json | allowedMcpServers | string | — | Enterprise allowlist undefined = allow all | deniedMcpServers | string | — | Enterprise denylist takes precedence over allow | allowManagedMcpServersOnly | boolean | false | Managed only Only admin-defined MCP servers apply | | Setting | Type | Default | What It Does | |---|---|---|---| hooks | object | {} | Custom commands before/after tool executions | disableAllHooks | boolean | false | Kill all hooks and statusLine execution | allowManagedHooksOnly | boolean | false | Managed only Block user/project/plugin hooks | allowedHttpHookUrls | string | — | URL patterns HTTP hooks may target supports wildcard | httpHookAllowedEnvVars | string | — | Env vars HTTP hooks may use in headers | | Setting | Type | Default | What It Does | |---|---|---|---| respectGitignore | boolean | true | @ file picker respects .gitignore patterns | autoMemoryEnabled | boolean | true | Auto-save useful context to .claude/memory/ | claudeMdExcludes | string | | Glob patterns for CLAUDE.md files to skip loading | cleanupPeriodDays | integer | — | Days to retain chat transcripts 0 = keep forever | plansDirectory | string | "~/.claude/plans" | Where plan files are stored | fileSuggestion | object | — | Custom script for @ file autocomplete | | Setting | Type | Default | What It Does | |---|---|---|---| apiKeyHelper | string | — | Script path that outputs auth values | awsCredentialExport | string | — | Script that exports AWS credentials | awsAuthRefresh | string | — | Script that refreshes AWS auth | forceLoginMethod | "claudeai" | "console" | — | Force login via Claude Pro/Max or Console billing | forceLoginOrgUUID | string | — | Organization UUID for OAuth login | otelHeadersHelper | string | — | Script that outputs OpenTelemetry headers | | Setting | Type | Default | What It Does | |---|---|---|---| includeGitInstructions | boolean | true | Include git commit/PR workflow instructions in system prompt | attribution | object | — | Customize attribution for git commits and PRs | autoUpdatesChannel | "stable" | "latest" | "latest" | stable = ~1 week old, skips regressions. latest = newest | skipWebFetchPreflight | boolean | false | Skip WebFetch blocklist check enterprise | sandbox | object | — | Sandbox execution configuration | teammateMode | "auto" | "tmux" | "iterm" | "in-process" | "auto" | How agent team teammates display | worktree | object | — | Config for --worktree sessions | env | object | {} | Environment variables for sessions many hidden settings live here | These go inside the "env" object and unlock settings not available as top-level properties: | Variable | What It Does | |---|---| CLAUDE CODE MAX TOOL USE CONCURRENCY | Max parallel tool calls default varies | CLAUDE CODE DISABLE AUTO MEMORY | 1 to disable auto memory | CLAUDE CODE DISABLE GIT INSTRUCTIONS | 1 to disable git instructions | CLAUDE CODE EFFORT LEVEL | Override effort level | CLAUDE CODE EXPERIMENTAL AGENT TEAMS | 1 to enable agent teams | CLAUDE CODE ENABLE TELEMETRY | 1 to enable telemetry | CLAUDE CODE DEBUG LOG LEVEL | verbose for debug logging | DISABLE AUTOUPDATER | 1 to disable auto-updates | ANTHROPIC MODEL | Runtime model override | CLAUDE CODE SUBAGENT MODEL | Model for subagents |