i18n-hunt
Find unused i18n translation keys in JavaScript projects using AST analysis.
✨ Why this exists
Managing i18n keys gets messy fast.
As applications evolve, translation files tend to accumulate stale keys that are no longer referenced anywhere in the codebase.
i18n-hunt helps identify those unused translations before they become noise.
🚧 Status
Experimental (WIP)
The project is currently in an early stage and under active development.
The goal is to validate the approach, gather feedback, and evolve it into a stable CLI.
🚀 Installation
npm install @jeevesinc/i18n-hunt -D
⚙️ Usage
Create an i18n-hunt.toml file:
locales = "./public/locales/en-US"
src = "./src"
# Optional
src_exclude = ["**/*.test.ts", "legacy/**"]
locales_exclude = ["Legacy/**"]
Add a script to your package.json:
{
"scripts": {
"hunt": "i18n-hunt"
}
}
Run the scan:
npm run hunt
You can also run it directly:
i18n-hunt --locales "public/locales/en-US" --src "src/"
🧠 Detection strategy
i18n-hunt analyzes your source code using AST analysis.
It classifies translation usage into:
- Static keys → directly detected (
t("form.email")) - Prefixes → partially dynamic but still safe (
t(form.${field})) - Dynamic usage → tracked without aggressively marking keys as unused
This helps reduce false positives while still surfacing genuinely unused translations.
For supported patterns and detection details, see the Wiki.
🤝 Contributing
Contributions and feedback are welcome — especially at this stage.
Useful contributions include:
- Reporting false positives or false negatives
- Sharing real-world edge cases
- Improving CLI UX
- Suggesting workflow or config improvements
