Detailed Analysis
A developer has released an open-source tool called aitrack, designed to aggregate and visualize AI coding assistant usage across multiple machines and platforms. The tool reads usage data from Claude Code, OpenAI's Codex, and Cursor, merging that data into unified heatmaps and statistical summaries. The project was motivated by the creator's personal need to track their own multi-device, multi-tool workflow without having to generate separate visualizations for each environment. The tool is available on GitHub and invites community contributions through feedback, feature requests, and bug reports.
One of aitrack's most notable design choices is its commitment to privacy and user control. Rather than relying on third-party servers or requiring account creation, it syncs data across machines through a user-controlled git repository. This approach ensures that no usage telemetry is transmitted to external services, positioning aitrack as a self-sovereign alternative to any analytics that might otherwise require trusting a vendor. In an environment where concerns about data privacy and AI tool monitoring are increasingly prominent, this design philosophy is likely to resonate strongly with developers who are cautious about exposing their workflow patterns or project details to outside parties.
Beyond the heatmap visualization, aitrack provides token statistics and cost estimates, addressing a practical concern for developers who use multiple AI coding assistants with consumption-based pricing. As AI tools like Claude Code and Codex bill based on token usage, having a consolidated view of consumption across platforms allows developers to make more informed decisions about which tools they rely on most and where their spending is concentrated. This kind of cross-platform cost visibility is not natively offered by any of the individual vendors, making aitrack a useful complement to existing tooling.
The emergence of tools like aitrack reflects a broader trend in the developer community toward meta-tooling — software built to manage, analyze, and optimize the use of AI development assistants themselves. As the AI coding assistant market has matured with competing offerings from Anthropic, OpenAI, and third-party IDEs like Cursor, developers increasingly find themselves working across multiple platforms simultaneously rather than committing to a single ecosystem. This fragmentation creates demand for aggregation layers that treat AI tools as a category of measurable infrastructure rather than isolated products. Aitrack occupies that niche, and its privacy-first, git-based architecture suggests a design sensibility attuned to the preferences of technically sophisticated, self-hosted-leaning users.
Read original article →