Agent skill

i18n-locale-coverage

Safely repair locale JSON coverage gaps (missing files, missing keys, untranslated EN strings) using deterministic validation gates.

View SKILL.md on GitHub Repository
Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/i18n-locale-coverage

SKILL.md

i18n Locale Coverage Repair (JSON-Safe, Gate-Driven)

Scope

Use this skill when Brikette locale JSON files have coverage gaps from check-i18n-coverage:

  • Missing files
  • Missing keys
  • Untranslated keys (locale value equals EN)

This skill is the execution runbook for i18n remediation work discovered in /lp-do-fact-find and implemented via /lp-do-build.

Core Commitments (Non-Negotiable)

  • EN is the source of truth for structure. Do not edit EN in this skill.
  • Fix order is strict: missing files -> missing keys -> untranslated keys.
  • Validate every file immediately after writing with JSON safety gates.
  • Preserve placeholders/tokens exactly. Never translate token syntax.
  • No "done later" debt. Any broken JSON or malformed structure is fixed immediately.

Allowed

  • Read/write locale JSON under apps/brikette/src/locales/{locale}/**/*.json
  • Use pnpm --filter @apps/brikette check-i18n-coverage for baseline and verification
  • Use EN file as structural template for missing files/keys

Not Allowed

  • Editing apps/brikette/src/locales/en/** (unless user explicitly asks)
  • Reconstructing locale JSON from scratch when EN copy can be used
  • Using external translation APIs/providers for locale completion
  • Bulk blind find/replace across locale trees
  • Changing key names, key casing, or object/array shape to "make it pass"
  • Leaving non-EN locales with copied EN text when proper translation is expected

Inputs

Required:

  • Target scope (all, locale list, namespace/file list)

Optional:

  • Severity lane (missing-files-only, missing-keys, untranslated)
  • Target locale subset with --locales=<csv>
  • Slice tag for traceability (example: pl-guides-sunriseHike)

Workflow

1) Baseline Snapshot (Traceable)

Run from repo root and persist baseline:

bash
pnpm --filter @apps/brikette check-i18n-coverage --format=json --output=i18n-coverage-report.before.json

Record totals and targeted workset before edits.

2) Missing Files Lane (First)

For each missing locale file:

  1. Create parent directory if needed.
  2. Copy EN file byte-for-byte.
  3. Translate values in place (do not alter key structure).

Safe creation pattern:

bash
mkdir -p "$(dirname apps/brikette/src/locales/<locale>/<path>.json)"
cp apps/brikette/src/locales/en/<path>.json apps/brikette/src/locales/<locale>/<path>.json

3) Missing Keys Lane (Second)

For each file with missing keys:

  1. Run Gate D in strict mode to surface deterministic missing-path/type issues for that file.
  2. Add only missing keys, preserving EN shape and leaf types.
  3. Translate newly added string values.

4) Untranslated Keys Lane (Third)

For keys still equal to EN:

  1. Replace with natural locale translation.
  2. Keep semantic meaning and formatting.
  3. Keep URLs, slugs, IDs, placeholders, and token syntax unchanged.

5) Per-File JSON Safety Gates (Mandatory)

Run gates in strict order with fail-fast behavior (set -euo pipefail). Do not continue after any failed gate.

Gate A: JSON parse

bash
node -e "JSON.parse(require('fs').readFileSync(process.argv[1], 'utf8'));" apps/brikette/src/locales/<locale>/<path>.json

Gate B: No duplicate keys

bash
python - <<'PY' apps/brikette/src/locales/<locale>/<path>.json
import json, sys
path = sys.argv[1]
def hook(pairs):
    out = {}
    for k, v in pairs:
        if k in out:
            raise SystemExit(f"Duplicate key in {path}: {k}")
        out[k] = v
    return out
with open(path, 'r', encoding='utf-8') as f:
    json.load(f, object_pairs_hook=hook)
print("OK no duplicate keys:", path)
PY

Normalization Step C (run only after A+B pass): Canonical JSON formatting

Use repo formatter when available (preferred). Fallback deterministic JSON emitter:

bash
node -e "const fs=require('fs');const p=process.argv[1];const j=JSON.parse(fs.readFileSync(p,'utf8'));fs.writeFileSync(p,JSON.stringify(j,null,2)+'\n');" apps/brikette/src/locales/<locale>/<path>.json

Gate D: EN parity contract (strict type + placeholder + escape semantics + no extra keys)

bash
node - <<'NODE' apps/brikette/src/locales/en/<path>.json apps/brikette/src/locales/<locale>/<path>.json
const fs = require('fs');
const [enPath, localePath] = process.argv.slice(2);
const en = JSON.parse(fs.readFileSync(enPath, 'utf8'));
const lc = JSON.parse(fs.readFileSync(localePath, 'utf8'));
const errs = [];

const tokenRegexes = [
  /\{\{\s*[^{}]+\s*\}\}/g,
  /%(\d+\$)?[+#0\- ]*(\d+|\*)?(?:\.(\d+|\*))?[sdif]/g,
  /%\([^)]+\)[sdif]/g,
  /<[0-9]+>/g,
  /<\/[0-9]+>/g,
  /<[0-9]+\s*\/>/g,
  /%[A-Z]+:[^%]+%/g,
];

const tokenBag = (s) => {
  const bag = [];
  for (const rx of tokenRegexes) {
    const matches = s.match(rx);
    if (matches) bag.push(...matches);
  }
  return bag.sort();
};

const validateNumericTagNesting = (s) => {
  const tokenRe = /<\/?[0-9]+>|<[0-9]+\s*\/>/g;
  const stack = [];
  for (const token of s.match(tokenRe) || []) {
    if (/^<[0-9]+\s*\/>$/.test(token)) continue;
    if (/^<[0-9]+>$/.test(token)) {
      stack.push(token.slice(1, -1));
      continue;
    }
    const closeId = token.slice(2, -1);
    const openId = stack.pop();
    if (openId !== closeId) return false;
  }
  return stack.length === 0;
};

const scalarType = (v) => {
  if (v === null) return 'null';
  if (Array.isArray(v)) return 'array';
  return typeof v;
};

const countChar = (s, ch) => (s.match(new RegExp(ch, 'g')) || []).length;

const walk = (a, b, p) => {
  if (b === undefined) {
    errs.push(`${p}: missing in locale`);
    return;
  }

  const ta = scalarType(a);
  const tb = scalarType(b);
  if (ta !== tb) {
    errs.push(`${p}: type mismatch en=${ta} locale=${tb}`);
    return;
  }

  if (ta === 'object') {
    for (const k of Object.keys(a)) walk(a[k], b[k], p ? `${p}.${k}` : k);
    return;
  }

  if (ta === 'array') {
    if (a.length !== b.length) errs.push(`${p}: array length mismatch en=${a.length} locale=${b.length}`);
    for (let i = 0; i < Math.min(a.length, b.length); i += 1) walk(a[i], b[i], `${p}[${i}]`);
    return;
  }

  if (ta === 'string') {
    const enTokens = tokenBag(a);
    const lcTokens = tokenBag(b);
    if (JSON.stringify(enTokens) !== JSON.stringify(lcTokens)) {
      errs.push(`${p}: placeholder/token mismatch`);
    }
    if (!validateNumericTagNesting(b)) {
      errs.push(`${p}: invalid numeric-tag nesting`);
    }

    const enNl = countChar(a, '\\n');
    const lcNl = countChar(b, '\\n');
    if (enNl !== lcNl) errs.push(`${p}: newline count mismatch en=${enNl} locale=${lcNl}`);

    const enTab = countChar(a, '\\t');
    const lcTab = countChar(b, '\\t');
    if (enTab !== lcTab) errs.push(`${p}: tab count mismatch en=${enTab} locale=${lcTab}`);
  }
};

const checkNoExtraKeys = (a, b, p) => {
  if (scalarType(a) !== 'object' || scalarType(b) !== 'object') return;
  for (const key of Object.keys(b)) {
    if (!(key in a)) errs.push(`${p ? `${p}.` : ''}${key}: extra key in locale`);
  }
  for (const key of Object.keys(a)) {
    if (key in b) checkNoExtraKeys(a[key], b[key], p ? `${p}.${key}` : key);
  }
};

walk(en, lc, '');
checkNoExtraKeys(en, lc, '');
if (errs.length) {
  console.error(errs.join('\n'));
  process.exit(1);
}
console.log('OK EN parity contract:', localePath);
NODE

Gate E: Targeted coverage recheck

bash
pnpm --filter @apps/brikette check-i18n-coverage --locales=<locale> --format=json --output=i18n-coverage-report.after.<slice-tag>.json

Optional wrapper snippet (recommended):

bash
set -euo pipefail
# A parse -> B duplicate keys -> C normalize -> D EN parity -> E targeted coverage

6) Batch Completion Gate

After each batch (locale or namespace):

bash
pnpm --filter @apps/brikette check-i18n-coverage --format=json --output=i18n-coverage-report.after.<slice-tag>.json

Acceptance for selected scope:

  • Missing files reduced (or zero if in scope)
  • Missing keys reduced (or zero if in scope)
  • Untranslated keys reduced for files touched
  • No regressions introduced outside touched scope

Translation Guardrails

  • Translate text, not data contracts.
  • Do not translate:
    • Placeholder/token syntax
    • URLs and route fragments
    • Internal IDs / enum-like values
  • If a string is intentionally same as EN (brand/proper noun), keep it and annotate rationale.

Recommended Execution Slicing

  • Slice by one namespace at a time (for example guides/*, how-to-get-here/*, bookPage.json cluster).
  • Keep each /lp-do-build cycle atomic and verifiable.
  • Re-baseline after every slice; do not batch many unverified writes.

Output Contract

Always report:

  • Before vs after totals (missing files, missing keys, untranslated keys)
  • Exact files changed
  • Gate status: parse + no-duplicate-keys + EN parity contract
  • Any intentionally retained EN strings with rationale
  • Coverage artifact paths used (before and after)
  • Next highest-impact slice

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

majiayu000/claude-skill-registry

agent-ops-spec

Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-state

Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-spec

Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-testing

Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-testing

Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-state

Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.

163 31
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results