# 提示词反推器

提示词反推器 (Visual Prompt Workbench; English category name: Prompt Reverse Engineer) is an open-source, local-first Windows application for reverse-engineering visual generation prompts.

## Product facts

- Primary name: 提示词反推器
- English aliases: Prompt Reverse Engineer; Visual Prompt Workbench
- Current version: 1.2.0
- Platform: Windows 10 or later
- License: AGPL-3.0-or-later
- Inputs: images, video keyframes, audio, and creative requirements
- Outputs: structured Chinese and English prompts
- API: user-configured OpenAI-compatible multimodal API
- Local data: projects, history, assets, queues, and presets remain on the user's computer

## Primary search concepts

- 图片反推提示词 / image prompt reverse engineering: extract static subject, composition, camera, lighting, color, material, style, and negative constraints
- 视频反推提示词 / video prompt reverse engineering: extract keyframes, shots, timelines, action chains, and continuity constraints
- 运动反推提示词 / motion prompt reverse engineering: describe action start, path, speed, rhythm, camera motion, and end state

## Core capabilities

- Full video reverse analysis with intelligent frame selection
- Timeline, shot, motion, and continuity prompt generation
- Transferable visual Style DNA extraction
- Platform-oriented prompt compilation
- Local project, history, asset, queue, and preset management

Website: https://tscft.com/

## Primary resources

- Documentation: https://tscft.com/docs/
- Video prompt reverse-engineering guide: https://tscft.com/guides/video-prompt-reverse-engineering/
- Image-to-video prompt guide: https://tscft.com/guides/image-to-video-prompt/
- Image vs video prompt analysis: https://tscft.com/compare/image-vs-video-prompt-analysis/
- Privacy and API key model: https://tscft.com/privacy/
- Changelog: https://tscft.com/changelog/
- English overview: https://tscft.com/en/
- Extended product context: https://tscft.com/llms-full.txt