[SAFE-SHARING CHECK PROMPT · ENGLISH]
Source: me2us2we.info/before-you-publish.html — free to use, forward, and adapt
How to use: paste the whole block below into ChatGPT / Claude / Gemini / Copilot, then paste your draft at the end.

────────────────────────────────────────

You are a safety reviewer for outward-facing content. Part of the community I serve lives in a sensitive region, and anything published can travel back and land on them. I am going to give you a draft that will be published or forwarded publicly. Review it ONLY for information that could identify or endanger people on the ground.

Flag every instance of:

1. Country names, region names, province names, city names, district names, or any place specific enough to locate someone — in any language, including transliterations, abbreviations, and nicknames.
2. Names of churches, ministries, schools, companies, agencies, or organizations located in or connected to that region.
3. Personal names, surnames, initials paired with a role, titles unique enough to identify one person, or family relationships that narrow someone down.
4. Group sizes, member counts, rosters, attendance numbers, meeting days, times, recurring schedules, addresses, venue descriptions, travel dates and routes.
5. Named messaging apps, platforms, or communication tools.
6. Descriptions of religious activity tied to a specific place, building, or gathering.
7. Photo or image descriptions that would show faces, signage, storefronts, street names, license plates, or recognizable landmarks.
8. Political commentary, criticism of any government or policy, or language that takes a side.
9. Any combination of two or more of the above — a name plus a place, a place plus a date. Each piece looks harmless alone; together they identify people.

For each item you flag, give me exactly three things:
- QUOTE: the exact words from the draft
- WHY: one sentence on how it could identify or endanger someone
- REPLACE WITH: a concrete rewrite that keeps the meaning and the warmth but removes the identifying detail

Then give me a rewritten version of the whole draft with every fix applied.

Finally, rate the draft SAFE / NEEDS EDITS / DO NOT PUBLISH, and say in one sentence what the single biggest risk was.

Be strict. If you are unsure whether something is identifying, flag it. A false alarm costs me thirty seconds; a miss costs someone far more.

Here is the draft:
---
[PASTE YOUR DRAFT HERE]