agentsclimarketplace

Lokalise core workflow b

Skill jeremylongshore/claude-code-plugins-plus-skills/plugins/saas-packs/lokalise-pack/skills/lokalise-core-workflow-b

425 plugins, 2,810 skills, 200 agents for Claude Code. Open-source marketplace at tonsofskills.com with the ccpi CLI package manager.

Install
npx -y skills add jeremylongshore/claude-code-plugins-plus-skills --skill lokalise-core-workflow-b

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

What its author says it does

Copied from the file, not written here

'Manage Lokalise secondary workflow: Download translations and integrate with app.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

12.4 KB, ~3.1k tokens by cl100k_base, as published. Nobody here has run it

Lokalise Core Workflow B

Overview

Everything on the "Lokalise to app" side: download translated files, manage translations and review status, leverage translation memory, manage contributors and their language access, and handle format differences across JSON, XLIFF, and PO files.

Prerequisites

  • Lokalise API token exported as LOKALISE_API_TOKEN
  • Lokalise project ID exported as LOKALISE_PROJECT_ID
  • @lokalise/node-api installed for SDK examples
  • lokalise2 CLI installed for CLI examples
  • unzip available for extracting download bundles

Instructions

  1. Download translated files. The download endpoint returns an S3 URL to a zip bundle — request the bundle, download the zip, then extract.

SDK — Download and extract:

import { LokaliseApi } from "@lokalise/node-api";
import { execSync } from "node:child_process";
import { mkdirSync } from "node:fs";

const client = new LokaliseApi({ apiKey: process.env.LOKALISE_API_TOKEN! });
const PROJECT_ID = process.env.LOKALISE_PROJECT_ID!;

// Request the download bundle
const download = await client.files().download(PROJECT_ID, {
  format: "json",
  original_filenames: false,
  bundle_structure: "%LANG_ISO%.json",     // Output: en.json, fr.json, de.json
  filter_langs: ["en", "fr", "de", "es"],  // Only these languages
  export_empty_as: "base",                 // Use base language for empty translations
  include_tags: ["release-3.0"],           // Only keys with this tag
  replace_breaks: false,
});

const bundleUrl = download.bundle_url;
console.log(`Bundle URL: ${bundleUrl}`);

// Download and extract
mkdirSync("./locales", { recursive: true });
execSync(`curl -sL "${bundleUrl}" -o /tmp/lokalise-bundle.zip`);
execSync(`unzip -o /tmp/lokalise-bundle.zip -d ./locales`);
console.log("Translations extracted to ./locales/");

CLI — Download with structure:

set -euo pipefail
lokalise2 --token "$LOKALISE_API_TOKEN" file download \
  --project-id "$LOKALISE_PROJECT_ID" \
  --format json \
  --original-filenames=false \
  --bundle-structure "locales/%LANG_ISO%.json" \
  --filter-langs "en,fr,de,es" \
  --export-empty-as base \
  --replace-breaks=false \
  --unzip-to .

SDK — Download with original file structure preserved:

const download = await client.files().download(PROJECT_ID, {
  format: "json",
  original_filenames: true,
  directory_prefix: "",                    // No extra prefix
  export_empty_as: "skip",                 // Omit untranslated keys
  include_comments: false,
  include_description: false,
});
  1. Manage translations — list, update, and mark as reviewed.

SDK — List translations for a language:

const frTranslations = await client.translations().list({
  project_id: PROJECT_ID,
  filter_lang_id: 673,            // Language ID for French (find via languages endpoint)
  filter_is_reviewed: 0,          // Only unreviewed
  limit: 100,
});

for (const t of frTranslations.items) {
  console.log(`[${t.key_id}] ${t.translation} (reviewed: ${t.is_reviewed})`);
}

SDK — Update a translation:

const updated = await client.translations().update(TRANSLATION_ID, {
  project_id: PROJECT_ID,
  translation: "Nouvelle traduction",
  is_reviewed: false,              // Mark as needing review after edit
});

SDK — Mark translations as reviewed (batch):

const unreviewed = await client.translations().list({
  project_id: PROJECT_ID,
  filter_lang_id: LANG_ID,
  filter_is_reviewed: 0,
  limit: 500,
});

for (const t of unreviewed.items) {
  await client.translations().update(t.translation_id, {
    project_id: PROJECT_ID,
    is_reviewed: true,
  });
}

console.log(`Marked ${unreviewed.items.length} translations as reviewed`);

SDK — List translations with cursor pagination (for large datasets):

async function* paginateTranslations(
  client: LokaliseApi,
  projectId: string,
  langId: number
) {
  let cursor: string | undefined;
  do {
    const params: Record<string, unknown> = {
      project_id: projectId,
      filter_lang_id: langId,
      limit: 500,
    };
    if (cursor) params.cursor = cursor;

    const page = await client.translations().list(params);
    yield* page.items;
    cursor = page.hasNextCursor() ? page.nextCursor() : undefined;
  } while (cursor);
}

// Usage
for await (const t of paginateTranslations(client, PROJECT_ID, 673)) {
  console.log(`${t.key_id}: ${t.translation}`);
}
  1. Leverage translation memory (TM) for auto-suggestions based on previously translated segments.

SDK — Use TM during upload:

const tmResults = await client.translationProviders().list({
  team_id: TEAM_ID,
});

// TM is automatically applied during file upload when `use_automations: true`
const upload = await client.files().upload(PROJECT_ID, {
  data: base64Data,
  filename: "en.json",
  lang_iso: "en",
  use_automations: true,       // Apply TM and MT suggestions automatically
  slashn_to_linebreak: true,
});

SDK — Leverage TM during download (pre-translate empty keys):

// Pre-translate uses TM + MT before download
// First, trigger pre-translation
// Then download with filled translations
const download = await client.files().download(PROJECT_ID, {
  format: "json",
  original_filenames: false,
  bundle_structure: "%LANG_ISO%.json",
  export_empty_as: "base",    // Fallback to base language if TM has no match
});
  1. Manage contributors — add translators and configure language access.

SDK — Add a translator with specific language access:

const contributor = await client.contributors().create({
  project_id: PROJECT_ID,
  contributors: [
    {
      email: "[email protected]",
      fullname: "Marie Dupont",
      is_admin: false,
      is_reviewer: true,
      languages: [
        {
          lang_iso: "fr",
          is_writable: true,       // Can edit French translations
        },
        {
          lang_iso: "de",
          is_writable: false,      // Read-only access to German
        },
      ],
    },
  ],
});

console.log(`Added contributor: ${contributor.items[0].email}`);

SDK — List all contributors:

const contributors = await client.contributors().list({
  project_id: PROJECT_ID,
  limit: 100,
});

for (const c of contributors.items) {
  const langs = c.languages.map(
    (l: { lang_iso: string; is_writable: boolean }) =>
      `${l.lang_iso}${l.is_writable ? "(rw)" : "(r)"}`
  ).join(", ");
  console.log(`${c.fullname} <${c.email}> — ${langs}`);
}

SDK — Update contributor permissions:

await client.contributors().update(CONTRIBUTOR_ID, {
  project_id: PROJECT_ID,
  is_reviewer: true,
  languages: [
    { lang_iso: "fr", is_writable: true },
    { lang_iso: "es", is_writable: true },   // Grant Spanish write access
  ],
});
  1. Handle file format differences across JSON flat, JSON nested, XLIFF, and PO.

JSON flat (react-i18next default):

{
  "greeting.hello": "Hello",
  "greeting.goodbye": "Goodbye",
  "errors.network": "Network error"
}

Download config:

const download = await client.files().download(PROJECT_ID, {
  format: "json",
  json_unescaped_slashes: true,
  original_filenames: false,
  bundle_structure: "%LANG_ISO%.json",
  placeholder_format: "icu",          // {name} style
  export_sort: "a_z",
});

JSON nested (next-intl, vue-i18n):

{
  "greeting": {
    "hello": "Hello",
    "goodbye": "Goodbye"
  },
  "errors": {
    "network": "Network error"
  }
}

Download config — use _ as key separator so Lokalise nests on .:

set -euo pipefail
lokalise2 --token "$LOKALISE_API_TOKEN" file download \
  --project-id "$LOKALISE_PROJECT_ID" \
  --format json \
  --original-filenames=false \
  --bundle-structure "%LANG_ISO%.json" \
  --export-key-name-as "key_name_dot_separated" \
  --unzip-to ./locales

XLIFF 1.2 (iOS, Angular):

const download = await client.files().download(PROJECT_ID, {
  format: "xliff",
  original_filenames: false,
  bundle_structure: "%LANG_ISO%.xliff",
  export_empty_as: "empty",
});

PO / GNU gettext:

set -euo pipefail
lokalise2 --token "$LOKALISE_API_TOKEN" file download \
  --project-id "$LOKALISE_PROJECT_ID" \
  --format po \
  --original-filenames=false \
  --bundle-structure "%LANG_ISO%/LC_MESSAGES/messages.po" \
  --unzip-to ./locales

Upload format auto-detection:

// Lokalise detects format from filename extension
// Just make sure the filename matches the content format
await client.files().upload(PROJECT_ID, {
  data: base64Data,
  filename: "messages.xliff",   // Triggers XLIFF parser
  lang_iso: "en",
});

Output

  • Downloaded translation files extracted to project directory
  • Translations updated and review status managed
  • Contributors added with appropriate language permissions
  • Files exported in the correct format for your i18n framework

Error Handling

ErrorCauseSolution
404 Project not foundWrong project_idRun client.projects().list() to verify
Empty bundle (0 files)No translations match filtersRemove include_tags / filter_langs to broaden
400 Invalid formatUnsupported export format stringUse: json, xliff, po, strings, xml, yaml
Download timeoutLarge project with many languagesFilter to specific languages with filter_langs
403 ForbiddenContributor lacks write accessCheck contributor language permissions
curl: (28) Operation timed outS3 bundle URL expired (valid ~30 min)Request a fresh download URL

Examples

Build-Time Translation Fetch

// scripts/fetch-translations.ts — run in CI before build
import { LokaliseApi } from "@lokalise/node-api";
import { execSync } from "node:child_process";
import { mkdirSync, readdirSync } from "node:fs";

const client = new LokaliseApi({ apiKey: process.env.LOKALISE_API_TOKEN! });
const PROJECT_ID = process.env.LOKALISE_PROJECT_ID!;

const download = await client.files().download(PROJECT_ID, {
  format: "json",
  original_filenames: false,
  bundle_structure: "%LANG_ISO%.json",
  export_empty_as: "base",
  export_sort: "a_z",
  replace_breaks: false,
});

mkdirSync("./src/locales", { recursive: true });
execSync(`curl -sL "${download.bundle_url}" -o /tmp/i18n.zip`);
execSync("unzip -o /tmp/i18n.zip -d ./src/locales");

const files = readdirSync("./src/locales").filter((f) => f.endsWith(".json"));
console.log(`Downloaded ${files.length} locale files: ${files.join(", ")}`);

Contributor Onboarding Script

const translators = [
  { email: "[email protected]", name: "Hans Mueller", langs: ["de"] },
  { email: "[email protected]", name: "Yuki Tanaka", langs: ["ja"] },
  { email: "[email protected]",  name: "Ana Garcia",  langs: ["es", "pt"] },
];

for (const t of translators) {
  await client.contributors().create({
    project_id: PROJECT_ID,
    contributors: [{
      email: t.email,
      fullname: t.name,
      is_admin: false,
      is_reviewer: false,
      languages: t.langs.map((l) => ({ lang_iso: l, is_writable: true })),
    }],
  });
  console.log(`Invited ${t.name} for ${t.langs.join(", ")}`);
}

Resources

Next Steps

For common errors and troubleshooting, see lokalise-common-errors. For production SDK patterns, see lokalise-sdk-patterns.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.