Claude Code Intelligence — Development skill for Claude Code
Local-first usage intelligence for Claude Code: an OTLP receiver, transcript parsing and quota tracking that answer where the tokens went and when the limit resets.
How to install Claude Code Intelligence
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open Furkiozknn/claude-code-intelligence and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Claude Code Intelligence does
Local-first usage intelligence for Claude Code: an OTLP receiver, transcript parsing and quota tracking that answer where the tokens went and when the limit resets. Costs are labelled estimates, never presented as billed amounts. Nothing leaves the machine. 256 tests.
Alternatives in Development
- Quosparr — 🐦 Native menu bar & tray app for macOS, Windows and Linux: track Codex, Claude Code and Grok quotas, resets 288 ★
- Tare — Diagnose Claude Code usage — where tokens went and why 256 ★
- Sc Estimate — Provide development estimates for tasks, features, or projects with intelligent analysis 55 ★
README

Çalışma adı: `claude-code-intelligence` (CLI `cci`, daemon `ccid` — geçici). Ad ürün kararıyla değişebilir.
AI coding agent kullanımını (önce Claude Code, mimari olarak diğer sağlayıcılar) **gözlemleyen, normalize eden, analiz eden, ölçen, tahmin eden, anomali tespit eden, uyaran ve "şu anda ne yapmalıyım?" sorusuna veriye dayalı cevap veren** yerel-öncelikli bir istihbarat/kontrol katmanı.
Bu bir usage monitor, token sayacı, kota görüntüleyici, pano, CLI aracı veya taskbar widget'ı **değildir** — bunların birleşiminden büyük bir sistem hedefleniyor.

Gerçek çıktı, elle yazılmış bir örnek değil: `python scripts/gorsel-uret.py` bu görseli `cci daily`'yi çalıştırıp üretir. Dikkat edilecek iki işaret — maliyetin yanındaki **≈** rakamın liste fiyatından türetilmiş bir tahmin olduğunu söyler, **`[withheld]`** ise satıcının kendi maliyetini vermediği yerde uydurulmadığını. Bu depoda tahmin, ölçümün yerine geçmez.
Yöntem ve durum
Research → Discover → Score → Deep Analyze → Extract → Synthesize
→ Design → Critique → Implement → Benchmark → Improve
✅ ✅ ✅ ✅ ✅ ✅
→ ✅ Design (Faz 7–9) → ✅ Implement (Stage 1–16)
**Kod durumu:** `cci/` paketi — model (Stage 1), adaptör sözleşmesi (2), toplayıcılar: zarf + ingest kapısı, transcript izleyici, kota poller, OTLP/JSON alıcı (3), SQLite olay deposu (4), normalizasyon + dedup/birleştirme (5), fiyat tablosu + özetler (6), pace v1 + boru hattı + snapshot + CLI (7).
uv sync && uv run pytest # 256 test
uv run cci scan # Claude Code + Codex transcript'lerini artımlı tara
uv run cci today | daily | sessions # özetler (--json, --strict → k
Related Skills
Auto Resume
Arm a local timer that types your next prompt into this terminal ~10 minutes after your rate limit resets. Arm
Claude Code Usage Quota Mod
Know your Claude limits before they hit you (Live band above the Claude Code prompt): 5 Hour and Weekly usage
Claude Quota Dash
A Claude Code quota dashboard on a LilyGO T-Display-S3. The host renders every pixel with Pillow; the board is
Codebase Graph
Code intelligence MCP server — 42-language tree-sitter AST parsing, FalkorDB knowledge graphs, 0.944 MRR searc
Burn Jira
Run one unattended pass of burn process jira over labelled Jira tickets — one session per ticket, stopping at
Claude Selfaware
Claude Code skill for automatic usage-limit awareness — checkpoints work and auto-resumes unattended after the
Related Agents
API Cost Scanner
Scans codebase for AI/API cost optimization opportunities including model usage, token estimates, caching stra
Cost Estimator
Estimates Cloudflare Workers costs before deploying. Reads wrangler.toml, counts D1 tables/rows, estimates KV/
10x Tool Calls
Cursor and Windsurf meter usage by requests and tool calls rather than tokens, which means a finished or stall