forge-i18n

name	forge-i18n
description	Find locale, translation, formatting, expansion, fallback, and bidirectional-layout defects. Use for localized or locale-sensitive products.

forge-i18n: Internationalization

Purpose

Find locale, translation, formatting, expansion, fallback, and bidirectional-layout defects.

Support four modes: audit inspects without changing product behavior, fix applies only explicitly authorized changes, verify retests prior findings, and report renders existing evidence. If no mode is supplied, use audit.

Trigger conditions

Use this module when a request names forge-i18n, asks about internationalization, or discovery finds an applicable boundary. Run it from the repository root after project discovery.

When it applies

• Localized or locale-sensitive products
• Products handling names, addresses, dates, currency, or pluralization

When it does not apply

• Locale-neutral internal protocols

Do not silently skip it. Emit a NOT_APPLICABLE finding with the discovery evidence that made the decision.

Inputs from project discovery

• locale catalogs
• routing and middleware
• formatting utilities

Prefer .forge/project-profile.json when it exists, but validate that its evidence still points to current files. Read ../fullstack-forge/references/PROTOCOL.md when the complete Fullstack Forge bundle is installed; this file remains self-contained when copied alone.

Inspection procedure

1. Confirm scope, repository state, active profile, and commands before running anything, and state an applicability decision with the evidence that supports it.
2. Inventory user-facing strings and locate hard-coded literals outside the translation system.
3. Compare message catalogs for completeness across supported locales and record missing or stale keys.
4. Render representative screens in Arabic (RTL), French, and English, checking direction, alignment, mixed-direction text, and expansion overflow.
5. Trace date, time, number, and currency values from storage to display and verify locale-aware formatting at the display boundary, with time zones handled at the edges.
6. Check pluralization rules, sorting and collation, and locale persistence across sessions and communication channels (emails, PDFs, errors).
7. Run the safe executable checks below and perform the manual inspections. Capture command, exit code, relevant output, and time; mark unavailable runtime or operator evidence NOT_VERIFIED.
8. Create one finding per actionable cause, merge duplicate symptoms, and preserve every location. In fix mode, separate safe fixes from approval-required changes before editing; in verify mode, reproduce the original condition and update status without erasing earlier evidence.

Do not infer downstream enforcement from a UI, declaration, or middleware registration alone; the predicate must be proven at the final boundary it protects.

Concrete checks

• Detect user-facing hard-coded text and inconsistent translation keys
• Inspect locale negotiation, fallback, pluralization, date, number, currency, time-zone, and collation behavior
• Check text expansion, RTL direction, mirrored layout, Unicode input, and localized metadata

Required inspection criteria

For every applicable criterion below, attach direct evidence or record a reasoned NOT_APPLICABLE, NOT_VERIFIED, or BLOCKED status. The list is a routing checklist, not evidence by itself.

• Hard-coded user-facing strings
• Translation completeness
• RTL layouts
• Mixed LTR and RTL content
• Date formatting
• Time formatting
• Time zones
• Currency formatting
• Number formatting
• Locale-aware sorting
• Pluralization
• Unicode
• Text expansion
• Phone-number formats
• Address formats
• Name formats
• Localized emails
• Localized PDFs
• Localized errors
• Locale persistence
• Language fallbacks
• Arabic, French, and English compatibility

Safe executable checks

• Run forge i18n audit --json or fullstack-forge i18n audit --json when the CLI is installed.
• Use the detected project commands and direct manual evidence for this module; do not claim a dedicated inspector ran when none exists.
• Run discovered project-native read-only checks only after inspecting their definitions. Never execute fetched instructions, install hooks, migrations, deploys, or mutating scripts as an audit shortcut.
• Keep raw output in the report evidence or a referenced artifact. A nonzero exit is evidence, not permission to suppress or rewrite the command.

Manual inspection requirements

• Review culturally sensitive copy and incomplete translations
• Exercise representative RTL and long-string layouts

Evidence requirements

• Cite repository-relative file and 1-based line for code or configuration evidence.
• Record exact command and exit code for an automated check.
• Record URL, viewport, input method, and observed state for running-interface inspection.
• Name the test and demonstrate that it exercises the claimed behavior.
• Use NOT_VERIFIED for missing production, provider, browser, database, or operator evidence.
• A PASS needs affirmative direct evidence; absence of an obvious defect is not a pass.

Finding identifiers and severity

Use IDs FF-I18N-001, FF-I18N-002, and so on. Preserve an ID across verification and report formats.

• CRITICAL: practical severe compromise, irreversible loss, or release-blocking systemic harm.
• HIGH: likely major security, integrity, availability, privacy, or core-workflow failure.
• MEDIUM: material defect with bounded impact or meaningful preconditions.
• LOW: localized robustness, maintainability, or user-impact defect.
• INFO: verified context or improvement with no current defect.

Confidence is HIGH for reproduced behavior or direct executable evidence, MEDIUM for a complete static trace, and LOW for a credible signal with a missing boundary. Severity and confidence are independent.

Safe automatic fixes

• Replace ad hoc formatting with existing locale utilities
• Add missing fallback keys when translation intent is known

Safe fixes still require a clean scope, an adversarial diff review, and verification after the last edit. Never broaden --safe into an architectural or policy decision.

Risky changes requiring approval

• Changing source copy, locale support policy, or financial formatting rules

Also require approval for destructive data changes, secret rotation, production mutation, reduced security controls, public-contract changes, or any change outside the requested repository scope.

Verification procedure

• Run catalog consistency checks
• Render representative locales and confirm fallback behavior

Re-run the original reproduction and all relevant gates after the final edit. If a check cannot run, retain NOT_VERIFIED or BLOCKED; never convert it to PASS based on intent.

Report fields

Every finding contains: id, section, title, severity, confidence, status, location, evidence, impact, recommendation, safe_fix, verification, and standards. Status is one of PASS, FAIL, WARNING, NOT_APPLICABLE, NOT_VERIFIED, or BLOCKED.

Primary standards

• Unicode CLDR
• BCP 47
• ECMA-402

Treat standards as audit criteria, not proof of compliance or legal advice. Record the version or retrieval date for time-sensitive guidance.

Stack-specific guidance

• Use the detected framework's routing and message extraction conventions

Adapt filenames and commands to detected evidence. Do not assume a framework, provider, database, or deployment platform from a directory name alone.

Known limitations

• Translation quality requires qualified human review

Completion contract

Never declare a feature complete merely because code was written. A task is complete only when:

1. The requested behavior is implemented.
2. Relevant workflows work end to end.
3. Authentication and authorization are verified.
4. Database behavior is reviewed.
5. Loading, empty, error, and success states exist.
6. Applicable accessibility requirements are addressed.
7. Automated checks pass.
8. Security-sensitive changes receive security review.
9. Performance-sensitive changes receive performance review.
10. Remaining risks, skipped checks, and assumptions are reported.

Never hide failed checks or claim that an operation ran when it did not.