lurx/smb-ai-skills-hub/israeli-social-content

Israeli Social Content Locale Layer

Overview

Localize strategy and copy for Israeli audiences on behalf of [שם העסק]. This skill does not choose topics, invent platform strategy, or schedule posts. It converts approved content into natural Hebrew and flags local calendar, RTL, and cultural risks.

Adapted from dangogit/content-skills (MIT — see LICENSE).

When to Use

• Rewrite approved copy into spoken Israeli Hebrew.
• Review mixed Hebrew, English, emoji, numbers, mentions, and hashtags.
• Check culturally sensitive dates or local context.
• Adapt one approved idea across Israeli-facing networks.

Prerequisites

• Approved idea, story role, and proof.
• The business CLAUDE.md (audience, brand positioning). If writing-style is

installed, its voice profile governs phrasing.

• Current official platform documentation for limits, labels, and formats

(claims expire — see references/evidence.json for dated claims).

Workflow

Step 1: Preserve Message

Keep original claim, proof, and promise. Do not add Israeli stereotypes, invented slang, unsupported statistics, or generic local references.

Step 2: Rewrite In The Business Voice

• Spoken, direct, second-person Hebrew.
• Prefer phrases the business actually uses (writing-style corpus, past posts).
• Start from lived pain before tool name.
• Reject translated slogan symmetry and invented framework names.
• Keep technical English only when needed.
• Isolate short Latin tokens on their own line or bidi-safe element.
• No inline AI in Hebrew. No em dashes.

Step 3: Validate RTL

• Start each Hebrew line with a Hebrew character.
• Put Latin hashtags and mentions on separate final lines.
• Use <bdi dir="ltr"> in HTML for code and commands.
• Use directional marks only when the target surface supports them reliably.
• Test in the actual composer or final rendered pixels before publishing.

Step 4: Check Cultural Timing

• Block insensitive commercial content on Yom Kippur and Yom HaZikaron unless

the user explicitly chooses another treatment.

• Check current Hebrew-calendar dates from a live source.
• Treat Shabbat, holidays, August, and election periods as hypotheses, not

universal engagement laws.

• Use the account's own performance history for timing, when available.

Step 5: Adapt Platform Surface

• Adapt only networks the user requested for this asset.
• Preserve one core promise while adapting title, caption, and media needs.
• Do not hardcode posting frequency, best time, ideal length, hashtag count,

or network mix.

• Do not claim Hebrew always outperforms English without account evidence.

Step 6: Check AI Disclosure

• TikTok requires labeling realistic AI-generated media; Meta requires

disclosure for photorealistic generated video/audio.

• Verify current official policy before production handoff.
• Do not claim compliant labels automatically reduce reach.

Run the mechanical copy gate before handoff:

python scripts/check_hebrew_copy.py <copy.md>

It catches em dashes, generic share CTAs, and inline AI mixed into Hebrew. It does not replace a native-speaker read, rendered RTL inspection, or live platform-policy verification.

Output Format

# Israeli Localization Review
 
Localized copy:
Voice check:
RTL risks:
Latin-token treatment:
Calendar risks:
Platform checks requiring live verification:
Unsupported claims removed:

Resources

• references/locale-checklist.md — concise localization and policy checklist.
• references/israeli-social-platforms.md — directional audience context with

methodology caveats.

• references/evidence.json — dated claim registry with confidence and expiry.

Related Skills

• writing-style — the business voice profile (use when installed).
• carousel — carousel production; run this skill as its copy gate.

Key Principles

1. Locale layer does not replace editorial strategy.
2. The business's spoken phrasing beats generic Israeli slang.
3. RTL correctness is a production requirement.
4. Account evidence beats hardcoded timing.
5. Platform policy claims expire and require verification.