---
name: tbt-annotated-image-comparison
description: "Ask AI for evidence-tied annotations rather than a generic judgement for Create an annotated before-and-after comparison. Use permitted source material; this kit does not authorise external actions."
---

# Create an annotated before-and-after comparison

Ask AI for evidence-tied annotations rather than a generic judgement.

Use the supplied source as authority. Check the account supports the actual file or image operation before attempting it.

## Task instruction

Build an editable before-and-after comparison from these permitted
actual files. The original controls [appearance/product facts/rules].
The candidate was asked to change [bounded task]. Use numbered regions
and separate requested improvements, visible source differences and
unverified details. Keep the images fixed and notes editable.
Do not invent outputs, measurements, approval or automatically detected
features. Include a plain-text note list and one targeted next request.

## Source boundaries

- Useful scene contact
- Unrequested plant
- Interpreted product geometry

## Review and correction

An annotation should name a visible region and a next action. It does not prove strict product fidelity.

Annotation [ID] identifies [specific visible source difference] at
[region]. Use the original as authority and correct that issue only,
keeping [useful requested improvement]. Do not add unrelated detail or
reshape the subject to make the repair easier. Return one actual file
for another comparison and keep unverified claims out of the note.

Use references/practice-source.json for the labelled example. The practice HTML is a local teaching tool; it does not call an AI service. Apply this method to the user's supplied facts rather than copying fictional example values. Do not infer approval or permission to publish.
