HduSy/tokenscope — MacOS menu-bar dashboard for Claude CLI token usage, estimated cost, and per-model / MCP / Skill bre…

Description and use

**Brief Summary (2‑3 sentences)**
HduSy /tokenscope is a Rust‑based macOS menu‑bar utility that visualises Claude CLI token consumption, estimated costs, and detailed per‑model, MCP, and Skill breakdowns. By surfacing these metrics in real time, it helps developers monitor and optimise the expense of AI‑driven workflows directly from the desktop. The project is actively maintained, well‑starred, and positioned as a lightweight bridge between Claude‑based agents and the Model Context Protocol (MCP) ecosystem.

**Value**

• **Cost transparency:** Shows exact token usage and projected monetary cost, preventing surprise bills when scaling AI assistants.
• **Model‑level insight:** Breaks down usage by individual Claude models, MCP endpoints, and custom Skills, enabling fine‑grained optimisation of prompts and routing logic.
• **Developer ergonomics:** A native menu‑bar UI means developers can watch token metrics without leaving their terminal or IDE, encouraging a “pay‑as‑you‑go” mindset.

**Practical Adoption Path**

1. **Install** the binary (or build from source) on any macOS workstation.
2. **Configure** the tool to point at the Claude CLI token (or API key) used by your agents.
3. **Enable** MCP or Skill endpoints in your Claude‑based services; tokenscope will automatically detect and label traffic.
4. **Iterate**: use the live dashboard to identify high‑cost models or over‑used Skills, then adjust prompt design, model selection, or MCP routing accordingly.
5. **Scale**: for teams, distribute the binary via internal package managers (Homebrew, Cargo) and standardise the configuration file across developers, ensuring consistent cost monitoring across all AI‑enabled projects.

**Production Readiness**

• **Activity & Community:** 207 ★, 11 forks, recent commits (last update 2026‑07‑05) and a focused Rust codebase indicate an active maintainer base.
• **Integration Simplicity:** Exposes a clear CLI/SDK interface and emits standard MCP metadata, making it straightforward to plug into existing Claude CLI pipelines.
• **Risk Profile:** No obvious licensing or security red flags, though a final audit of the repository’s license compliance and supply‑chain hygiene is advisable.
• **Pilot Suitability:** Given its low runtime footprint, native macOS UI, and real‑time metrics, tokenscope is ready for a serious pilot in production environments where cost control and model observability are critical.

Problem

HduSy/tokenscope helps connect AI assistants to real tools and data through a standard protocol.

Outcome

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

Workflow

• connect AI agents to tools
• ship Model Context Protocol servers
• standardize integrations

Evidence

MacOS menu-bar dashboard for Claude CLI token usage, estimated cost, and per-model / MCP / Skill breakdowns.

Source