# Feedback For Learning

> Give feedback that improves performance, timed and framed so it is used rather than defended against. Use when coaching, reviewing work, or mentoring.

- Skill: `amey-thakur/feedback-for-learning` (Agent Skill)
- Install (CLI): `npx skillmds@latest add amey-thakur/feedback-for-learning`
- Raw SKILL.md: https://api.skillmd.com/api/skills/amey-thakur/feedback-for-learning/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: Amey-Thakur (https://skillmd.com/u/amey-thakur)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/amey-thakur/feedback-for-learning

---


# Feedback for learning

Feedback that is accurate and badly delivered produces defence rather
than change. Effective feedback is specific, timely, about the work, and
paired with what to do next.

## Method

1. **Be specific about the work.** This function does two things and
   the name says one is actionable; this is unclear is not (see
   giving-feedback).
2. **Deliver it close to the action.** Feedback arriving weeks later
   cannot be connected to the decisions that produced the work.
3. **Separate the observation from the judgement.** What you saw, then
   its effect, then the suggestion, which keeps it discussable rather
   than a verdict.
4. **Give one or two things to change.** A long list is overwhelming and
   produces no change; the most important thing produces one.
5. **Explain the reasoning.** Feedback with a why transfers judgement,
   where a rule transfers only compliance (see mentoring-engineers).
6. **Ask before telling where the learner can find it.** A question that
   leads them to the problem teaches more than the answer does.
7. **Follow up on whether it helped.** Feedback nobody checks is
   advice, and the follow-up is what makes it coaching.

## Boundaries

Feedback requires enough trust to be heard, which precedes technique.
Feedback on people rather than work damages the relationship and changes
nothing. Performance conversations with employment consequences are a
different activity with different rules (see agent-people-ops-desk).