---
name: tbt-ai-feedback-that-teaches
description: "Require evidence, reasoning and a next attempt rather than generic praise or grading for Ask AI for feedback that explains a specific mistake. Use permitted source material; this kit does not authorise external actions."
---

# Ask AI for feedback that explains a specific mistake

Require evidence, reasoning and a next attempt rather than generic praise or grading.

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

## Task instruction

Review this draft against the supplied source and task.
For each material mistake, quote the affected draft phrase, point to
the source passage, explain the changed meaning and propose a correction.
Prioritise the errors that affect the reader's decision.
Explain one useful technique I can apply on the next attempt.
Do not give generic praise, a numerical grade or invented source evidence.
Keep sound parts of the draft rather than rewriting everything.

## Source boundaries

- Proposed 15 October
- Subject to approval
- Flawed draft says launches

## Review and correction

Explain the changed meaning and repair that cause. A generic score or nicer tone does not teach the next attempt.

The draft changed 'proposed 15 October, subject to approval' into
'launches 15 October'. Restore both the proposed status and approval
condition. Preserve the useful short format. Show the corrected sentence
and the source link, then check other dates for the same compression error.
Do not invent who will approve it or say approval has occurred.

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.
