# Clip Feedback

> Analyze an AI-generated video clip against the prompt that produced it. Detects canon rule violations, emotion/energy/world mismatches, palette drift, and quality issues, then suggests specific prompt fixes to improve the next generation. Use whenever the user has a Veo, Seedance, Kling, Runway, or other AI-generated clip and asks "why does this look wrong", "check the clip", "check this clip", "prompt feedback", "analyze this clip", "compare clip to prompt", or anytime a video was generated from a prompt and needs iteration. Trigger this skill even when the user casually asks "is this good" about an AI clip — prompt vs reality analysis is the job.

- Skill: `luishiluy/clip-feedback` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add luishiluy/clip-feedback`
- Raw SKILL.md: https://api.skillmd.com/api/skills/luishiluy/clip-feedback/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: LuisHiluy (https://skillmd.com/u/luishiluy)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/luishiluy/clip-feedback

---


# Clip Feedback — Prompt Intent vs Reality

## What this does

Takes a video clip plus the prompt that generated it, analyzes what the clip actually looks like (via Gemini 2.5 Pro), and returns:

- **Rule violations** — character names in prompt, 4th wall breaks, missing focal length, safety trigger stacking
- **Intent gaps** — emotion mismatch, world palette drift, shot type wrong, energy level off
- **Quality issues** — text overlays, CGI artifacts, motion blur, missing face
- **Specific prompt fixes** — exact corrections to improve the next generation

## Usage

### Quick rule check (no API call — instant)

```bash
python3 scripts/prompt_feedback.py \
  --clip /path/to/clip.mp4 \
  --prompt "the generation prompt used" \
  --rules-only
```

### Full analysis (calls Gemini 2.5 Pro — ~10s)

```bash
python3 scripts/prompt_feedback.py \
  --clip /path/to/clip.mp4 \
  --prompt "the generation prompt used" \
  --source veo
```

Supported sources: `veo`, `seedance`, `kling`, `runway`, or any label.

### From a prompt file

```bash
python3 scripts/prompt_feedback.py \
  --clip /path/to/clip.mp4 \
  --prompt-file /path/to/prompt.txt \
  --source seedance
```

### JSON output (for piping to other tools)

```bash
python3 scripts/prompt_feedback.py \
  --clip /path/to/clip.mp4 \
  --prompt "..." \
  --json
```

## When the user says "check the clip"

1. Ask which clip (path) if not obvious from context
2. Ask for or find the prompt that generated it
3. Run `--rules-only` first for instant feedback
4. If the user wants deeper analysis, run the full version
5. Present the feedback summary and suggested fixes
6. Offer to apply the fixes and regenerate

## What it checks (canonical rules)

| Rule | What it catches |
|---|---|
| No character names | "Jane" → should be "the daughter" |
| No 4th wall | "looking at camera" → eyeline off-camera |
| No resolution text | "2K" in prompt → handled by tool settings |
| No music/dialogue | "music" in prompt → add to avoid line |
| Safety stacking | 3+ trigger words in one sentence → split |
| Focal length required | Missing "85mm, f/1.4" → add lens spec |
| Material descriptors | No textures → add tactile nouns |
| World palette | Palette drift → strengthen palette direction |
| Emotion match | Prompt says "dread", clip reads "calm" → add atmosphere cues |
| Quality flags | Text overlays, CGI glow, motion blur → add negative prompts |

## Configuration

Configure your canonical rules in `reference/prompt-rules.md`. Each rule gets:
- A pattern (regex or keyword list) to detect violations in prompts
- A fix template (the correction to suggest)
- A severity (block / warn / info)

## Dependencies

- Python 3.10+
- `google-generativeai` for Gemini analysis
- `ffmpeg` on PATH for frame extraction

## Setup

```bash
pip install google-generativeai pillow
export GEMINI_API_KEY=your_key_here
```