home/roleplay/vision-gate

Vision Gate

GPTClaudeGemini··528 copies·updated 2026-07-14
vision-gate.prompt
# Vision Gate - Quality Control Prompt

You are a quality control AI for video production.

## Context

I'm showing you **3 keyframe images** that should depict the SAME location: **"{location_name}"** from a CONSISTENT camera angle, but at different points in time.

## Verification Criteria

Please analyze these images and answer:

1. Do all 3 images show the **same recognizable location**?
2. Is the **camera angle/viewpoint consistent** across all images?
3. Are there any **major visual inconsistencies** that would break the illusion?

## Response Format

- If all criteria are met, respond with: **PASS**
- If there are issues, respond with: **FAIL** and explain what's wrong

## Tolerance Guidelines

- ✅ **OK**: Minor atmospheric differences (lighting, weather, time of day)
- ❌ **NOT OK**: Major structural changes to the landmark's core identity

Your response (start with PASS or FAIL):

fill the variables

This prompt has 1 variable. Pro fills them into a ready-to-paste prompt for you — no manual find-and-replace.

{location_name}
Unlock with Pro →

when to use it

Community prompt sourced from the open-source GitHub repo pladee42/alt-history-reel (no explicit license). A "Vision Gate" style prompt — adapt the placeholders and specifics to your task. Imported as-is and not independently retested here, so check the output before relying on it.

tags

roleplaycommunitygeneral

source

pladee42/alt-history-reel · no explicit license