lokalise-prod-checklist
Checks translation coverage and configuration before Lokalise production launches.
Install
mkdir -p .claude/skills/lokalise-prod-checklist && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/8780" && unzip -o skill.zip -d .claude/skills/lokalise-prod-checklist && rm skill.zipInstalls to .claude/skills/lokalise-prod-checklist
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Execute Lokalise production deployment checklist and rollback procedures.Key capabilities
- →Audit translation coverage against thresholds
- →Detect missing keys between source code and Lokalise
- →Validate translation file formats
- →Verify API token security and configuration
- →Check fallback language settings
How it works
The skill runs automated scripts to compare translation progress, sync keys between code and the platform, and validate JSON file structures before production deployment.
Inputs & outputs
When to use lokalise-prod-checklist
- →Auditing translation coverage before release
- →Validating production API configuration
- →Verifying fallback language settings
- →Checking production deployment readiness
About this skill
Lokalise Production Checklist
Overview
A structured pre-deployment checklist for Lokalise integrations covering nine verification areas: translation coverage, missing key detection, format validation, API token security, rate limit preparedness, fallback language configuration, download verification, OTA configuration, and contributor access review. Run through each section before any production deployment.
Prerequisites
- Lokalise project with production API token
lokalise2CLI installed and authenticatedcurlandjqavailable in your environment- Access to the Lokalise dashboard (Team Owner or Admin role)
- Application codebase with i18n integration ready for deployment
Instructions
Step 1: Translation Coverage Audit
Verify that every supported locale meets the coverage threshold before deploying.
#!/bin/bash
# scripts/audit-translation-coverage.sh
set -euo pipefail
: "${LOKALISE_API_TOKEN:?Required}"
: "${LOKALISE_PROJECT_ID:?Required}"
REQUIRED_COVERAGE=100 # Percentage required for production
REQUIRED_LOCALES=("en" "de" "fr" "es" "ja")
echo "=== Translation Coverage Audit ==="
# Get project statistics from Lokalise API
STATS=$(curl -sf "https://api.lokalise.com/api2/projects/${LOKALISE_PROJECT_ID}/statistics" \
-H "X-Api-Token: ${LOKALISE_API_TOKEN}")
TOTAL_KEYS=$(echo "$STATS" | jq '.project_statistics.keys_total')
echo "Total keys in project: $TOTAL_KEYS"
# Get per-language statistics
LANGUAGES=$(curl -sf "https://api.lokalise.com/api2/projects/${LOKALISE_PROJECT_ID}/languages" \
-H "X-Api-Token: ${LOKALISE_API_TOKEN}")
FAILED=0
echo ""
echo "Locale | Progress | Words | Status"
echo "---------|----------|-----------|-------"
for locale in "${REQUIRED_LOCALES[@]}"; do
progress=$(echo "$LANGUAGES" | jq -r ".languages[] | select(.lang_iso == \"${locale}\") | .statistics.progress")
words=$(echo "$LANGUAGES" | jq -r ".languages[] | select(.lang_iso == \"${locale}\") | .statistics.words_to_do")
if [[ -z "$progress" || "$progress" == "null" ]]; then
echo "${locale} | MISSING | - | FAIL"
FAILED=1
continue
fi
if (( $(echo "$progress < $REQUIRED_COVERAGE" | bc -l) )); then
echo "${locale} | ${progress}% | ${words} remaining | FAIL"
FAILED=1
else
echo "${locale} | ${progress}% | 0 | PASS"
fi
done
echo ""
if [[ $FAILED -eq 1 ]]; then
echo "RESULT: FAILED — Resolve incomplete translations before deploying."
exit 1
fi
echo "RESULT: PASSED — All locales meet ${REQUIRED_COVERAGE}% coverage."
Step 2: Missing Key Detection
Detect keys present in source code but absent from Lokalise, and vice versa.
#!/bin/bash
# scripts/detect-missing-keys.sh
set -euo pipefail
: "${LOKALISE_API_TOKEN:?Required}"
: "${LOKALISE_PROJECT_ID:?Required}"
echo "=== Missing Key Detection ==="
# Download current translation keys from Lokalise
TEMP_DIR=$(mktemp -d)
trap 'rm -rf "$TEMP_DIR"' EXIT
lokalise2 file download \
--token "$LOKALISE_API_TOKEN" \
--project-id "$LOKALISE_PROJECT_ID" \
--format json \
--original-filenames=true \
--directory-prefix="" \
--unzip-to "$TEMP_DIR/" 2>/dev/null
# Extract keys from Lokalise download
LOKALISE_KEYS=$(jq -r '[paths(scalars)] | map(join(".")) | .[]' "$TEMP_DIR/en.json" | sort)
# Extract keys referenced in source code (adjust pattern for your framework)
# React/i18next: t('key'), t("key"), <Trans i18nKey="key">
SOURCE_KEYS=$(grep -roh "t(['\"][^'\"]*['\"])" src/ 2>/dev/null \
| sed "s/t(['\"]//;s/['\"])//" \
| sort -u || true)
echo ""
echo "Keys in Lokalise: $(echo "$LOKALISE_KEYS" | wc -l)"
echo "Keys in source code: $(echo "$SOURCE_KEYS" | wc -l)"
# Keys in code but not in Lokalise
MISSING_IN_LOKALISE=$(comm -23 <(echo "$SOURCE_KEYS") <(echo "$LOKALISE_KEYS") || true)
if [[ -n "$MISSING_IN_LOKALISE" ]]; then
echo ""
echo "MISSING IN LOKALISE (used in code but not uploaded):"
echo "$MISSING_IN_LOKALISE" | head -20
COUNT=$(echo "$MISSING_IN_LOKALISE" | wc -l)
[[ $COUNT -gt 20 ]] && echo " ... and $((COUNT - 20)) more"
fi
# Keys in Lokalise but not in code (potential dead keys)
ORPHANED=$(comm -13 <(echo "$SOURCE_KEYS") <(echo "$LOKALISE_KEYS") || true)
if [[ -n "$ORPHANED" ]]; then
echo ""
echo "ORPHANED IN LOKALISE (not referenced in code):"
echo "$ORPHANED" | head -20
COUNT=$(echo "$ORPHANED" | wc -l)
[[ $COUNT -gt 20 ]] && echo " ... and $((COUNT - 20)) more"
fi
if [[ -z "$MISSING_IN_LOKALISE" && -z "$ORPHANED" ]]; then
echo ""
echo "RESULT: PASSED — Keys are in sync."
fi
Step 3: Format Validation
Validate that downloaded translation files are well-formed JSON and contain no placeholder mismatches.
// scripts/validate-translation-format.ts
import fs from 'fs';
import path from 'path';
const LOCALES_DIR = 'src/locales';
const SOURCE_LOCALE = 'en';
const PLACEHOLDER_REGEX = /\{\{?\w+\}?\}|%[sd@]|\$\{[\w.]+\}/g;
interface ValidationResult {
locale: string;
valid: boolean;
errors: string[];
}
function validate(): ValidationResult[] {
const results: ValidationResult[] = [];
const sourceFile = path.join(LOCALES_DIR, `${SOURCE_LOCALE}.json`);
const sourceContent = JSON.parse(fs.readFileSync(sourceFile, 'utf-8'));
const sourcePlaceholders = extractPlaceholders(sourceContent);
for (const file of fs.readdirSync(LOCALES_DIR)) {
if (!file.endsWith('.json')) continue;
const locale = file.replace('.json', '');
const errors: string[] = [];
// Check 1: Valid JSON
let content: Record<string, unknown>;
try {
content = JSON.parse(fs.readFileSync(path.join(LOCALES_DIR, file), 'utf-8'));
} catch (e) {
errors.push(`Invalid JSON: ${(e as Error).message}`);
results.push({ locale, valid: false, errors });
continue;
}
// Check 2: No empty string values (should be null or absent)
const emptyKeys = findEmptyValues(content);
if (emptyKeys.length > 0) {
errors.push(`Empty string values: ${emptyKeys.slice(0, 5).join(', ')}${emptyKeys.length > 5 ? ` (+${emptyKeys.length - 5} more)` : ''}`);
}
// Check 3: Placeholder consistency
const localePlaceholders = extractPlaceholders(content);
for (const [key, expected] of Object.entries(sourcePlaceholders)) {
const actual = localePlaceholders[key];
if (actual && actual.sort().join(',') !== expected.sort().join(',')) {
errors.push(`Placeholder mismatch in "${key}": expected [${expected}], got [${actual}]`);
}
}
results.push({ locale, valid: errors.length === 0, errors });
}
return results;
}
function extractPlaceholders(obj: Record<string, unknown>, prefix = ''): Record<string, string[]> {
const result: Record<string, string[]> = {};
for (const [key, value] of Object.entries(obj)) {
const fullKey = prefix ? `${prefix}.${key}` : key;
if (typeof value === 'string') {
const matches = value.match(PLACEHOLDER_REGEX);
if (matches) result[fullKey] = matches;
} else if (typeof value === 'object' && value !== null) {
Object.assign(result, extractPlaceholders(value as Record<string, unknown>, fullKey));
}
}
return result;
}
function findEmptyValues(obj: Record<string, unknown>, prefix = ''): string[] {
const result: string[] = [];
for (const [key, value] of Object.entries(obj)) {
const fullKey = prefix ? `${prefix}.${key}` : key;
if (value === '') result.push(fullKey);
else if (typeof value === 'object' && value !== null) {
result.push(...findEmptyValues(value as Record<string, unknown>, fullKey));
}
}
return result;
}
// Run validation
const results = validate();
let failed = false;
for (const r of results) {
const status = r.valid ? 'PASS' : 'FAIL';
console.log(`[${status}] ${r.locale}`);
r.errors.forEach(e => console.log(` ${e}`));
if (!r.valid) failed = true;
}
process.exit(failed ? 1 : 0);
Step 4: API Token Security
echo "=== API Token Security Audit ==="
# Verify token is not hardcoded in source
HARDCODED=$(grep -r "X-Api-Token" --include="*.ts" --include="*.js" --include="*.json" \
-l src/ 2>/dev/null | grep -v node_modules || true)
if [[ -n "$HARDCODED" ]]; then
echo "FAIL: API token may be hardcoded in: $HARDCODED"
exit 1
fi
# Verify token is not in version control
GIT_SECRETS=$(git log --all -p --diff-filter=A -- '*.env' '*.env.*' 2>/dev/null \
| grep -i "LOKALISE_API_TOKEN=" | head -5 || true)
if [[ -n "$GIT_SECRETS" ]]; then
echo "FAIL: Token found in git history. Rotate immediately."
exit 1
fi
# Verify .env files are gitignored
if ! grep -q "\.env" .gitignore 2>/dev/null; then
echo "WARN: .env not in .gitignore"
fi
# Verify token permissions are minimal
TOKEN_RESPONSE=$(curl -sf "https://api.lokalise.com/api2/projects/${LOKALISE_PROJECT_ID}" \
-H "X-Api-Token: ${LOKALISE_API_TOKEN}" -o /dev/null -w "%{http_code}")
if [[ "$TOKEN_RESPONSE" == "200" ]]; then
echo "PASS: Token is valid and has project access"
else
echo "FAIL: Token returned HTTP $TOKEN_RESPONSE"
exit 1
fi
echo "PASS: No hardcoded tokens detected"
Step 5: Rate Limit Preparedness
echo "=== Rate Limit Readiness ==="
# Lokalise enforces 6 requests/second per token
# Check if your code handles 429 responses
RATE_LIMIT_HANDLING=$(grep -r "429\|rate.limit\|retry-after\|rateLimitRetry" \
--include="*.ts" --include="*.js" src/ 2>/dev/null | head -5 || true)
if [[ -z "$RATE_LIMIT_HANDLING" ]]; then
echo "WARN: No rate limit handling detected in source code."
echo " Add retry logic with exponential backoff for 429 responses."
echo " Lokalise limit: 6 requests/second per API token."
else
echo "PASS: Rate limit handling detected"
echo "$RATE_LIMIT_HANDLING"
fi
Step 6: Fallback Language Configuration
Verify the application gracefully falls back when a translation is missing.
// Verify in your i18n configuration:
// i18next example
import i18next from 'i1
---
*Content truncated.*
When not to use it
- →Deploying without verifying translation coverage
- →Using hardcoded API tokens in source code
Prerequisites
Limitations
- →Requires manual resolution of incomplete translations
- →Placeholder mismatch detection is limited to specific regex patterns
How it compares
This provides a structured, automated checklist for production readiness rather than relying on manual verification of translation files.
Compared to similar skills
lokalise-prod-checklist side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| lokalise-prod-checklist (this skill) | 0 | 27d | Caution | Intermediate |
| build-macos-apps | 70 | 8mo | Review | Intermediate |
| cloudflare-manager | 25 | 9mo | Review | Intermediate |
| railway-cli-management | 9 | 8mo | Review | Intermediate |
Try saying
Example prompts that trigger this skill in your AI assistant.
More by jeremylongshore
View all by jeremylongshore →You might also like
build-macos-apps
glittercowboy
Build professional native macOS apps in Swift with SwiftUI and AppKit. Full lifecycle - build, debug, test, optimize, ship. CLI-only, no Xcode.
cloudflare-manager
qdhenry
Comprehensive Cloudflare account management for deploying Workers, KV Storage, R2, Pages, DNS, and Routes. Use when deploying cloudflare services, managing worker containers, configuring KV/R2 storage, or setting up DNS/routing. Requires CLOUDFLARE_API_KEY in .env and Bun runtime with dependencies installed.
railway-cli-management
CaptainCrouton89
Deploy, manage services, view logs, and configure Railway infrastructure. Use when deploying to Railway, managing environment variables, viewing deployment logs, scaling services, or managing volumes.
unity-editor-toolkit
Dev-GOM
Automate and control Unity Editor with 500+ commands, real-time WebSocket communication, and SQLite integration for efficient game development.
miniprogram-development
TencentCloudBase
WeChat Mini Program development rules. Use this skill when developing WeChat mini programs, integrating CloudBase capabilities, and deploying mini program projects.
azure-functions
aj-geddes
Create serverless functions on Azure with triggers, bindings, authentication, and monitoring. Use for event-driven computing without managing infrastructure.