androidZzT/cc-statistics — AI Coding stats dashboard — track costs, tokens, and efficiency across Claude Code / Gemini CLI / Co…

Description and use

**Brief Summary (2‑3 sentences)**
androidZzT/cc‑statistics is an open‑source dashboard that aggregates AI‑coding usage data—costs, token consumption, and efficiency—from Claude Code, Gemini CLI, Codex, and Cursor. It surfaces implementation signals such as API/SDK calls, language metadata, and topic focus, making it easy to prototype, evaluate, and monitor AI‑enhanced features. With active maintenance, 107 ⭐ on GitHub and recent updates, it’s a production‑ready candidate for teams that want AI capabilities without building a model stack from scratch.

**Value**

• **Unified visibility**: Consolidates telemetry from multiple LLM providers into a single UI, letting developers instantly see spend, token usage, and latency trends.
• **Rapid prototyping**: By exposing signals (API/SDK usage, language tags, topic focus) developers can quickly iterate on RAG pipelines, agent workflows, or other AI‑augmented features without writing custom instrumentation.
• **Cost control & efficiency**: Real‑time dashboards help teams spot waste (e.g., over‑tokenized prompts) and optimize prompt design, directly reducing operational expenses.

**Practical Adoption Path**

1. **Clone & install** – The project is Swift‑based; add it as a Swift Package or integrate the compiled binary into your iOS/macOS app or a cross‑platform UI layer.
2. **Configure data sources** – Provide API keys or CLI endpoints for Claude, Gemini, Codex, and Cursor in the supplied config.yaml. The dashboard auto‑discovers calls made through the official SDKs/CLIs.
3. **Instrument custom code (optional)** – If you have bespoke AI wrappers, emit the same telemetry format (JSON payload with model, tokens, cost, language, topic). The built‑in collector will ingest it.
4. **Deploy** – Run the dashboard locally for development, then containerize it (Dockerfile is included) for staging or production environments.
5. **Iterate & monitor** – Use the UI to set alerts on cost thresholds, compare provider efficiency, and guide prompt‑engineering decisions.

**Production Readiness**

• **Activity & community**: 107 stars, 14 forks, recent commit on 2026‑06‑28, and a healthy issue/PR turnover indicate an active maintainer base.
• **Stability**: The Swift codebase is mature, with clear module boundaries and a documented configuration schema.
• **Ecosystem fit**: Works out‑of‑the‑box with the major LLM providers used in modern dev‑tooling stacks, and the telemetry format aligns with common observability pipelines (e.g., Prometheus, Grafana).
• **Risks to address**: Final due‑diligence on licensing (MIT‑compatible) and a security audit of the data‑collection layer are recommended before a full production rollout.

Overall, cc‑statistics offers a low‑friction way to add AI observability and cost management to any project, making it a solid OSS candidate for pilot programs and, after the minor risk checks, for production use.

Problem

androidZzT/cc-statistics helps add AI capability without starting from a blank model stack.

Outcome

High for an OSS candidate: recent activity, adoption, and ecosystem signals are strong enough for a serious pilot.

Workflow

• prototype AI features
• build RAG or agent workflows
• evaluate model tooling

Evidence

AI Coding stats dashboard — track costs, tokens, and efficiency across Claude Code / Gemini CLI / Codex / Cursor

Source