i18n-LE: The Locale That Only Looks Finished

Audit translation catalogues against the source locale, in one keystroke
i18next, next-intl, VS Code bundles and Flutter ARB — keys and tokens, never a translated string

Useful? A star or rating is how other developers find it — ★ GitHub · ★ Open VSX · ★ Marketplace

What it does

Spanish shipped last week and the metrics screen has said {{periodo}} to every user since. The catalogue had every key. It parsed. The placeholder came back from machine translation with its name translated too, which compiles perfectly and renders the literal.

Open a catalogue, run i18n-LE: Audit Catalogues, and the set it belongs to is audited against the source locale: keys a locale is missing or has too many of, placeholders dropped or renamed in translation, constructs from another library's convention, values left empty, keys defined twice, and a path that is an object in one locale and a string in another. The report opens beside the editor. From the Explorer, run it on any folder. Works in VS Code and in VS Code–based editors like Cursor and VSCodium (installable from Open VSX).

• Before a release — is this locale shippable, or does it only look like it is?
• After machine translation — {{timeframe}} that came back as {{periodo}}, caught by its token names
• Reviewing a translator's pull request — the keys that changed shape, without reading the translations

The report never contains a translated string. It names keys, placeholder tokens, counts and shapes; an untranslated string is proved by being byte-identical to the source, not by being shown. It rewrites nothing.

Install

It works out which library you use first

{name} is a placeholder in next-intl and literal text in i18next; {{ name }} is an i18next variable and an ICU literal brace around a word. No amount of looking at the bytes settles that, so the library is identified before a catalogue is read, and the placeholder grammar, the plural model and which keys are metadata all come from the answer.

Five kinds of evidence, and two agreeing is an identification: a manifest dependency (package.json, pubspec.yaml), a config file, the directory layout, the catalogue's own syntax, and call sites in source. An ARB file's @@locale beside @key metadata settles it alone. When nothing agrees, or two libraries do, it refuses and names what it found rather than picking one. Set i18n-le.library to skip identification.

A VS Code extension's package.nls.json sits at the root while its translations live elsewhere; that split is read as one set. Source files are read only to answer which library this is, never for a finding.

What it finds

One catalogue is the contract, never the union of all of them — a union would turn one translator's typo into a missing key for everyone else. It is the one English catalogue, or the one i18n-le.source names. Plural categories are folded onto their base key, so Polish's legitimate _few is not an extra key; plural-category completeness is deliberately not checked.

Use it from an AI agent

The same engine runs as an MCP server, so an agent can call it directly instead of being handed twenty-five catalogues to diff itself.

check_catalogues(library, files, source?, keysAreSource?, maxResults?)

The tool takes catalogue contents and the library's name — it cannot see the project, so it cannot identify the library and does not guess. It returns the same report the editor renders, as data, capped at 500 findings by default with meta.truncated. It reads no files and makes no network requests. Published as i18n-le-mcp on npm and as io.github.nolindnaidoo/i18n-le in the MCP registry. It answers exactly as the Rust CLI's server does: one corpus runs against both, and a differential test feeds both thousands of generated catalogue sets — broken ones included, where both report the parser's own words and position — and compares every answer.

{
  "mcpServers": {
    "i18n-le": {
      "command": "npx",
      "args": ["-y", "i18n-le-mcp"]
    }
  }
}

echo '{"jsonrpc":"2.0","id":1,"method":"tools/list"}' | npx -y i18n-le-mcp

The CLI

The same audit runs from a terminal or a CI step: a Rust CLI in crate/, sharing one corpus with the extension — crate/fixtures/ — so the two can never read a catalogue differently.

i18n-le locales/                       # the set in that directory, as one JSON report
i18n-le --system vscode-l10n l10n/     # name the library, skip identification
i18n-le --fail-on untranslated locales/
i18n-le mcp                            # the same audit over MCP on stdio

Exit codes are the API — 0 clean, 1 findings, 2 the question was malformed (no library identified, two identified, no single source). Finding no catalogues at all is 0: there is nothing to be wrong with.

Commands

No command is bound to a key by default. Give any of them one under Keyboard Shortcuts in the editor.

Settings

Languages

Twelve languages besides English:

German · Spanish · French · Indonesian · Italian · Japanese · Korean · Portuguese (Brazil) · Russian · Ukrainian · Vietnamese · Chinese (Simplified)

Both halves are covered — the manifest (command titles, setting names and descriptions) and everything shown while the extension runs (notifications, the status bar and the report's headings). A refusal's reason is the engine's English, identical to the CLI's.

Privacy & security

• No network access. The extension never sends data anywhere. The telemetryEnabled setting only writes events to a local Output Channel you can inspect (i18n-LE).
• No translated value ever reaches a report, a notification or the MCP boundary. Keys are the deliberate exception, and where a layout makes the key the English sentence, that English is shown.
• The MCP server holds the same line. It takes content as an argument and returns data: no filesystem access, no network calls, no telemetry. check:mcp-bundle fails the build if a translation reaches its stdout.
• Error notifications redact home directories and credential-shaped fragments.
• One rating prompt, at most twice. On the 3rd successful use the extension asks once whether you would rate it, and once more on the 20th if you chose Later or dismissed it. Don't Ask Again ends it. Setting notificationsLevel to important or silent yourself turns it off. The counts are kept in VS Code's extension storage and nothing is sent anywhere; Rate opens the listing you installed from — the VS Code Marketplace or Open VSX — in your browser.

Documentation

Performance

Median of 7 runs after warmup, on Apple M5 Pro, 24 GB RAM, Node 24.3.0. Inputs are generated by scripts/benchmark.ts rather than checked in, so the sizes above are exactly what was measured. Reproduce with bun run benchmark.

These are machine-specific and are not asserted in CI — a benchmark that gates a build only tells you how busy the runner was.

Testing

244 test cases across 15 files, plus an integration suite that runs in a real VS Code extension host and an end-to-end test that installs the built .vsix into a clean profile.

Generated from a real run — coverage/coverage-summary.json and coverage/test-results.json — by scripts/coverage-readme.js; CI fails if this section drifts. Reproduce with bun run test:coverage, and the case count is the one vitest prints.

More from the LE family

Sixteen single-purpose tools for the work in front of every model. Each ships a Rust CLI and an MCP server. One page: letools.dev

Get it out

• String-LE — Extract every string in a codebase, with its position, so a person can read them
• Numbers-LE — Extract every hardcoded number in a codebase, so a person can check them
• Units-LE — Extract every quantity with its unit, normalized, and refuse the ambiguous ones by name
• Dates-LE — Extract every date and timestamp, and the exact instant each one resolves to
• IDs-LE — Extract every UUID, ULID, NanoID, ObjectId and Snowflake, and decode the time inside
• IPs-LE — Extract every IP address, CIDR block and MAC, normalized and classified by scope
• URLs-LE — Extract every URL in a codebase, with its protocol and exact position
• Paths-LE — Extract every file path in a codebase, and say whether it still points at anything
• Colors-LE — Extract every color in a codebase, and say which ones are not in your palette

Check it

• Regex-LE — Find every regex in a codebase, and report which can be driven into catastrophic backtracking
• Versions-LE — Find where one dependency is constrained differently across a repository's manifests
• i18n-LE — Identify the i18n library a project uses, then audit its catalogs by that library's rules
• Scrape-LE — Check whether a page is scrapeable before the scraper is written, and say when it cannot tell

Guard it

• Secrets-LE — Find hardcoded credentials in a codebase, and never print one into the report
• EnvSync-LE — Compare the dotenv files in a tree, and say which keys are missing from which
• Unicode-LE — Find the Unicode that hides meaning — bidi controls, invisibles, homoglyphs, mixed scripts

Each stands on its own: no shared crate, no published core. Where two of them agree, it is because the same answer was right twice.

Contact — nolindnaidoo.com · GitHub · LinkedIn

Also by nolindnaidoo

Rust — pixelcoords and pixelactions are one loop: pixelcoords answers where, pixelactions acts there. Their own tools, their own voice — not part of the LE family.

• pixelcoords — Freeze your screen, mark regions, get pixel-exact coordinates and crops pixelcoords.dev · crates.io · docs.rs
• pixelactions — Consume human-verified coordinates, perform the interaction, confirm it landed pixelactions.dev · crates.io · docs.rs

License

MIT © nolindnaidoo