---
id: oxbshw/watch-skill/learning-from-mistakes
version: "5f78ff24"
license: MIT
install: manual
updated: 2026-07-12
---

# learning-from-mistakes — When a video answer goes wrong, report the correction and Watch Skill records it as a lesson stored locally. The system classifies the mistake, injects it into similar future questions, and can re-ask the original question to verify the fix worked. Check what's been learned or review token savings anytime.

Publisher: oxbshw · Stars: 236 · Updated: 2026-07-12

Install (manual): `git clone https://github.com/oxbshw/watch-skill`

## SKILL.md

# Learning from mistakes

When a video answer turns out wrong, the correction is worth more than
an apology. Report it and Watch Skill learns locally — the mistake is
classified, stored as a lesson, injected into future similar questions,
and where possible the original question is immediately re-asked to
confirm the lesson actually fixes it.

## Report a correction

The moment the user corrects a video answer, run:

```bash
watch-skill lessons add <video_id> "<question>" "<the wrong answer>" "<the correction>"
```

The output includes the lesson (`error_class`, `guidance`) and, when the
question was re-asked, whether the corrected answer now comes back —
tell the user which it was. Nothing is uploaded; lessons live in
`~/.watch-skill/`.

## See what has been learned

```bash
watch-skill lessons list
```

## The savings meter

```bash
watch-skill stats
```

Lifetime token economics: what text-first answering and the semantic
cache have saved versus stuffing raw frames into context. Quote real
numbers from it when the user asks whether any of this is worth it.


[View on SkillFed](https://skillfed.io/oxbshw/watch-skill/learning-from-mistakes) · [View on GitHub](https://github.com/oxbshw/watch-skill)