Yuhan Lei
9340c1f581
Tokenizer rewritten in closed form (strict thousands grouping, dotted tokens as a whole) and frozen: quotes, ISO dates, and specific number forms are the documented candidate set; new prose forms extend the docstring, not the regexes. Quote pairing no longer skips short pairs into phantom quotes. Raw links parse only from the fenced-stripped header. Skip-files applies only to root-level index/log. Body blockquotes join the quote channel. Fence openers are indent-limited. Examples now honor their own invariant: two fabricated quotes restored to the raw wording, the market-data date moved into the raw body, and the index example rewritten as a real global-index excerpt. A smoke test assembles the examples and demands zero suspects, so the samples stay under their own contract. Archive pages get a defined index date (their Archived field) for lint cross-checks.
5.3 KiB
Claude Code Statusline Tool Ecosystem
Sources: claude-pace project research, 2026-03-19; 2026-03-24 Raw: raw/ai-coding-tools/2026-03-19-claude-code-statusline-landscape.md Updated: 2026-03-24
Overview
Around Claude Code's opaque quota problem, the community has spawned over a dozen statusline tools, forming a rapidly evolving ecosystem. By mid-March 2026, competitive focus shifted from "feature count" to "data accuracy and runtime reliability", with zero runtime dependencies becoming a differentiation point.
Competitive Landscape (2026-03-19 Data)
| Project | Stars | Language | Positioning |
|---|---|---|---|
| ccusage | 11,693 | TypeScript | CLI usage analysis tool, statusline is a sub-feature; core selling point: cost visualization |
| claude-hud | 7,038 | JavaScript | Most feature-complete pioneer: context progress bar, rate limit, sub-agent status, Git integration |
| Claude-Code-Usage-Monitor | 7,009 | Python | Standalone dashboard, ML prediction; discontinued (last update 2025-09) |
| ccstatusline | 5,421 | TypeScript | Positioned as "formatter", no API calls; highly customizable + themes |
| CCometixLine | 2,227 | Rust | High-performance binary; only Rust solution in awesome-claude-code |
| claude-powerline | 931 | TypeScript | vim powerline style |
Primary distribution channels:
- awesome-claude-code (29,014 stars): currently includes 5 statusline tools, excludes claude-hud
- Anthropic official plugin directory (12,653 stars): zero statusline entries
User Pain Points
Opaque Quota (Root Problem)
- Sudden limit hit without warning: $200/month Max users hit limits in 10-15 minutes, only see "usage limit reached"
- Invisible quota reset time: On 2026-03-18, 3 independent issues requested this feature on the same day
- 5h + 7d dual window cognitive burden: Users can't distinguish which limit they're under
Officially Not Doing
Anthropic marked issue #10593 as "Not Planned" (2026-01-19), rejecting native token indicator, recommending ccusage. Meanwhile, statusline stdin JSON doesn't expose session_used_percentage, weekly_used_percentage, resets_at quota fields, forcing third-party tools to use Usage API or parse local JSONL files.
User Feedback Analysis
Best-Received Features
Context progress bar: Multiple independent sources consistently call it "install reason". SAP community author: "context bar alone is worth the install".
Cross-session daily cost summary: ccusage's core selling point, Japanese user community extensively documents "one command to see costs" experience. Multiple independent reviews list "cost visualization" as primary reason for choosing ccusage.
Rate Limit 5h/7d visualization: After v2.1.80, stdin provides official data, all tools have equal opportunity, differentiation space shifts from "having it" to "being accurate".
Pitfalls Already Hit
ccusage Live Blocks removed: Real-time token consumption monitoring was officially removed due to accuracy issues (issue #782). Issues #288 (16 reactions) and #483 (11 reactions) document persistent user complaints about "shows not at limit but actually hit".
ccusage process management bug (issue #459, 10 reactions): bun x ccusage statusline in hooks causes infinite process spawning, CPU 100%. This is structural risk from Node.js runtime.
claude-hud sub-agent monitoring: Blog authors highly praise it, but 6 subagent-related issues all have 0 reactions, suspected maintainer or AI bulk creation, not user-driven. This shows systematic bias between blog review perspective (feature completeness) and user voting (most painful needs).
Feature Creep Warning
Ovidiu (Substack): "When everything is highlighted, nothing is"
ccusage's Live Blocks from feature to removal全过程 is a live lesson: real-time features with inadequate accuracy反而 damage tool credibility.
Technical Trends
Migration from Node.js to lightweight runtimes:
- Rust: CCometixLine, claudia-statusline
- Go: felipeelias (single binary, minimal)
- Shell/Bash: kamranahmedse, rz1989s (curl one-line install)
Zero-dependency install as selling point: Multiple community authors list "reduce dependencies" as tool selection factor, Go single binary and pure Bash solutions' motivation both include this.
No public performance benchmarks: No tool has published hyperfine or equivalent ms-level benchmark numbers, this is unoccupied differentiation space.
Impact
Statusline track in March 2026 saw rapid growth (claude-hud gained 4,804 stars in 5 days, mainly GitHub Trending effect), but the track has become crowded. Real competitive advantage doesn't come from feature count, but from:
- Data reliability: ccusage's live blocks removed due to inaccuracies, shows "accurate" is more important than "rich"
- Zero dependencies: Node.js tools' process management bugs show runtime dependency is structural risk, not just installation inconvenience
- Differentiated features: Unique core concepts (like pace tracking: is current speed enough) are more competitive than "me too" features
Features not worth pursuing: Sub-agent monitoring (data source not in stdin, actual user demand evidence insufficient), Multi-device data sync (high engineering complexity, weak demand evidence).