agentleFS
opencollectiveopencollective/opencollective-frontend/.agents/skills/i18n-frontend-translate/SKILL.md

i18n-frontend-translate

Skill897 starsChanged 52 days agoMITView on GitHub

Batches translation of locale JSON using show-untranslated and set-translation scripts, with codebase lookup for message IDs when context is unclear. Use when the user asks to translate frontend strings, reduce untranslated entries for a locale, or apply translation rules to lang files.

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 4 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    List work: Run show-untranslated with a small batch so the task stays reviewable:

    line 22
  2. Very specific:

    Read the line: {id}: {english}.

    line 31
  3. Distinctive:

    If meaning, audience, or UI role is unclear, find context before translating (see below).

    line 32
  4. Very specific:

    Update progress: After finishing (or when the user stops), refresh locale completion percentages in lib/constants/locales.js:

    line 46
  5. Distinctive:

    Use when the English line alone is ambiguous (button vs heading, legal vs casual, technical sense).

    line 58
  6. Very specific:

    Search the repo for the id (appears in FormattedMessage, defineMessages, intl.formatMessage, etc.):

    line 60
  7. Very specific:

    Keep ICU placeholders and rich-text tags exactly as in English unless the target language requires a different order: {name}, {count}, …

    line 76
  8. Distinctive:

    Do not invent new placeholders or remove required tags.

    line 77
  9. Very specific:

    Prefer consistent product terminology with the rest of lang/<locale>.json when the same concept appears elsewhere.

    line 78

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

i18n Frontend Translate

When to Use

  • User gives a locale (e.g. fr, pt-BR) and optionally instructions.
  • Goal is to replace locale strings that still equal English with proper translations, using the repo scripts.

Preconditions

  • Work from this repository root.
  • lang/<locale>.json must exist. English source of truth is lang/en.json.

Workflow (loop)

  1. User inputs: locale + free-text instructions (follow them for register, formality, product terms, and consistency).
  2. List work: Run show-untranslated with a small batch so the task stays reviewable:

    npx tsx scripts/i18n/show-untranslated.ts <locale> --limit 30

    Increase --limit or re-run without --limit when draining the backlog.

  3. Per string:

    • Read the line: {id}: {english}.
    • If meaning, audience, or UI role is unclear, find context before translating (see below).
    • If the string should stay English in this locale (same word in both languages, standard loanword, product name, etc.), do not run set-translation. Add the message id to IGNORED['<locale>'] in scripts/i18n/translation-stats.ts so it stops appearing in the report. Keep a short inline comment per entry (e.g. why it is not translated). Ignore matching is by id, not by English value.
    • Otherwise produce the target-language string that matches instructions and preserves placeholders/markup, then apply:
    npx tsx scripts/i18n/set-translation.ts <locale> <id> "Translation text"

    Quote the translation so the shell does not split words.

  4. Repeat step 2 until show-untranslated returns nothing (or the user stops).
  5. Sanity check: Optionally re-run show-untranslated for that locale; remaining lines should be actual translation work, or items intentionally listed under IGNORED.
  6. Update progress: After finishing (or when the user stops), refresh locale completion percentages in lib/constants/locales.js:

    npm run langs:update-progress

Intentionally identical strings (IGNORED)

show-untranslated reports ids whose locale value still equals English, minus IGNORED for that locale. When a value is correctly left as English, register the message id under IGNORED in scripts/i18n/translation-stats.ts. Otherwise the loop never clears.

Finding context for a message id

Use when the English line alone is ambiguous (button vs heading, legal vs casual, technical sense).

  • Search the repo for the id (appears in FormattedMessage, defineMessages, intl.formatMessage, etc.):

    rg -l '<id>' --glob '*.{tsx,ts,js,jsx}'

    Example: rg -l 'Ri4REE' --glob '*.{tsx,ts,js,jsx}'

  • Open the matching file(s) and read surrounding UI: labels, modals, routes, audience.
  • For duplicate or generic English, context disambiguates (e.g. "Order" as noun vs verb).
  • Compare with other languages or en.json by searching lang/.

Preservation rules

  • Keep ICU placeholders and rich-text tags exactly as in English unless the target language requires a different order: {name}, {count}, {amount}, <Link>...</Link>, <Account></Account>, etc.
  • Do not invent new placeholders or remove required tags.
  • Prefer consistent product terminology with the rest of lang/<locale>.json when the same concept appears elsewhere.

Scripts reference

Script / npm scriptRole
scripts/i18n/show-untranslated.tsLists {id}: english where locale value still equals English, minus per-locale IGNORED message ids. Optional --limit N.
scripts/i18n/set-translation.tsWrites one key: set-translation.ts <locale> <id> "text"
npm run langs:update-progressRecomputes per-locale completion % and writes lib/constants/locales.js. Run once when done.

Show-untranslated and set-translation are invoked with npx tsx from the repository root.

Download now (ZIP)

  • In this repositoryOne other file it gives its agents.

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.