- Home
- Skills
- Translation & Localization
- Czech ↔ English i18n
Works with the AI Agents you already use
Czech ↔ English i18n
Correct Czech UI translations — three plural forms, gender, formal vy — with a checker that catches raw keys before users do.
Free
Czech ↔ English i18n
Example session with this skill installed
Check our Czech translation file before release: src/locales/cs.json against en.json.
- Read your context and instructions
- Compiled the czech english i18n
scripts/i18n_check.py src/locales/cs.json src/locales/en.json (real run on a production manufacturing app)
8 errors, 5 warnings, 68 info — exit code 1, so CI would block the release.
| Key | Problem | What users actually saw (rendered with the app's own i18next 25) |
|---|---|---|
vyroba.stats.recordCount | _one/_few/_many, no _other | 5 records → the raw key vyroba.stats.recordCount |
smrsteni.spc.inWarningZone | _one/_other, no _few | 3 points → the raw key smrsteni.spc.inWarningZone |
fifo.group.lotCount | old lotCount + _other only | 3 → "3 lot" instead of "3 loty" |
vykaz.form.savedCount | placeholder {{suffix}} only in Czech | the strings disagree on inputs — check the code passes suffix |
Why
In i18next v21+ Czech _many is only for decimals (1,5). Integers 0 and 5+ use _other, 2–4 use _few.
Fix
"recordCount_one": "{{count}} záznam",
"recordCount_few": "{{count}} záznamy",
"recordCount_many": "{{count}} záznamu",
"recordCount_other": "{{count}} záznamů"
Re-run the checker until it exits 0, then look at the changed screens in Czech — long Czech words break narrow buttons.
Connects securely to your tools. The creator never sees your data.
About this skill
Translation tools and LLMs often fail at the nuances of Czech grammar, particularly complex pluralization and gender agreement. This skill provides a specialized workflow for developers managing i18n between English and Czech, ensuring production-ready translations that follow linguistic rules and technical standards.
What it does
- Validates plurals by ensuring all three Czech forms (1, 2–4, 5+) are present for count-based strings.
- Enforces grammar including gender agreement for past-tense verbs and adjectives based on the subject noun.
- Checks consistency via a bundled Python script that detects missing keys, duplicates, and placeholder mismatches.
- Format compliance for numbers using comma decimals, spaces in dates, and 24-hour time conventions.
- Framework detection to apply the correct syntax for i18next, ICU, vue-i18n, next-intl, or gettext.
How it works
- Analyze environment to identify the localization library and existing key naming conventions.
- Execute validation using
i18n_check.pyto find broken plural sets or missing keys incs.jsonanden.json. - Generate translations that apply sentence casing, mandatory diacritics, and appropriate formal or informal tone.
- Audit placeholders to ensure they remain in the nominative case or are phrased to avoid declension errors.
Frameworks & tools
Supports i18next, react-i18next, vue-i18n, next-intl, ICU MessageFormat, and gettext. Includes a Python-based validation script for CI/CD integration.
Why this beats prompting it yourself
Generic LLMs often use English plural logic or incorrect suffixes like _many for integers, which causes raw keys to render in production. This skill encodes specific Czech linguistic rules and library-specific implementation quirks that general-purpose prompts miss.
Use cases
- Localizing a React or Vue application from English to Czech for the first time.
- Fixing "missing plural" bugs where 5 or more items show the raw translation key.
- Auditing existing locale files for duplicate keys and untranslated strings.
- Standardizing UI tone between formal "Vy" and informal "Ty" across a large codebase.
Known limitations
Does not support automatic translation of complex domain-specific legal or medical terminology without a provided glossary. Requires Python 3 to run the validation script.
Requires Python 3 for the checker script. Does not translate complex domain-specific terminology without context. Focuses on i18next, ICU, vue-i18n, next-intl, and gettext formats.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
2 uses
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 1 month ago
- Free with an Agensi account
Requires
Permissions
Terminal runs the read-only checker (scripts/i18n_check.py). Read Files: your locale files. Write Files: only when you ask the agent to add or fix translations. No network access.