How do I use i18n Doctor to check translation health?
4 मिनट पठनमध्यम
i18n Doctor analyzes your project's translation health and gives you a score out of 100. It checks for missing translations, unused keys, inconsistencies, and other common issues.
Accessing Doctor
There are two ways to run health checks:
From the dashboard
- Go to your project → Integrations → Doctor
- Click Quick Health Check
- Review the results
From CI/CD
Add Doctor to your CI pipeline to catch issues before merge:
# GitHub Actions
- name: Check translation health
run: better-i18n doctor --project acme/dashboard
Doctor score
Your score is shown in the project header as a badge (e.g., "Doctor 85/100"). The score is calculated from:
| Check | Weight | What it measures |
|---|---|---|
| Missing translations | High | Keys without translations in target languages |
| Coverage balance | Medium | Whether all languages have similar coverage |
| Unused keys | Low | Keys in dashboard but not in code |
| Orphaned keys | Low | Keys in code but not in dashboard |
Understanding results
Doctor reports show:
- Score — overall health (0-100)
- Issues — list of problems found, grouped by severity
- Recommendations — actionable suggestions to improve your score
Severity levels
| Severity | Meaning |
|---|---|
| Critical | Missing translations in production languages |
| Warning | Low coverage, potential unused keys |
| Info | Suggestions for improvement |
Setting up CI integration
For continuous monitoring:
- Go to Integrations → Doctor
- If you have a GitHub repo connected, click Set up CI
- Follow the instructions to add the Doctor workflow to your repository
- Doctor will run on every pull request and report results
Report history
Previous reports are stored and accessible from the Doctor page. Click any report to see its full details. Use the ?reportId= URL parameter to deep-link to a specific report.
Tips for a high score
- Translate all keys before publishing — aim for 100% coverage per language
- Remove unused keys — keep your project clean
- Run Doctor regularly — catch issues early
- Set up CI checks — prevent regressions
Next steps
क्या यह लेख सहायक था?