Skip to content

Latest commit

 

History

22 Commits

Folders and files

Repository files navigation

daftAI-skills

English | 中文

Skills shared by daftAI for improving daily work efficiency with AI Agent.

Prerequisites

  • Node.js environment installed
  • Ability to run npx bun commands

Installation

Quick Install (Recommended)

npx skills add daftAI2026/daftAI-skills

Available Skills

Skill Description
daftai-video-subtitler Burn subtitles into local videos. Supports SRT/VTT/ASS formats, bilingual subtitle merging and translation
daftai-subtitle-translator Translate subtitle files (SRT/VTT) between languages with terminology consistency
daftai-extracting-video-subtitles Extract timed SRT subtitles from video/audio files using OpenAI Whisper
daftai-url-to-markdown Fetch any URL and convert to markdown using Chrome CDP, with automatic image downloading
daftai-chinese-copywriting Check and fix Chinese copywriting (punctuation, spacing, full-width/half-width) using autocorrect
daftai-image-to-svg Convert transparent or solid-background single-color PNG/SVG artwork into measured, path-only SVG output
daftai-gpt-image-2-5-prompt-optimizer Optimize natural-language requests into GPT Image 2.5 prompts using live official guidance; returns text, not images

Usage

Each skill has its own SKILL.md with detailed instructions. The AI assistant will automatically follow these instructions when the skill is loaded.

Example: Extract Subtitles

User: Extract subtitles from this video
→ AI loads daftai-extracting-video-subtitles skill
→ bun scripts/extract_subtitles.ts video.mp4 en
→ Output: video.srt

Example: Burn Subtitles

User: Burn Chinese subtitles into this video
→ AI loads daftai-video-subtitler skill
→ Detects environment, scans subtitles, burns to video
→ Output: video_zh.mp4

Example: Optimize an Image Prompt

User: Use daftai-gpt-image-2-5-prompt-optimizer: a watercolor postcard of an orange cat in a rainy alley, no text.
→ If `references/gpt-image-2-5/` already matches the published layout, AI reads the local index; otherwise it fetches once
→ Later runs reuse that local official reference, with no image API key

Project Structure

daftAI-skills/
├── README.md
├── README.zh.md
├── CHANGELOG.md
├── CHANGELOG.zh.md
├── AGENTS.md
├── .gitignore
└── skills/
    ├── daftai-video-subtitler/
    │   ├── SKILL.md
    │   ├── scripts/          # TypeScript utilities
    │   └── references/       # Config docs
    ├── daftai-subtitle-translator/
    │   ├── SKILL.md
    │   └── references/       # Glossary
    ├── daftai-extracting-video-subtitles/
    │   ├── SKILL.md
    │   └── scripts/          # TypeScript utilities
    ├── daftai-url-to-markdown/
    │   ├── SKILL.md
    │   └── scripts/          # TypeScript utilities
    ├── daftai-chinese-copywriting/
    │   ├── SKILL.md
    │   ├── scripts/          # TypeScript utilities
    │   └── references/       # Config docs
    ├── daftai-image-to-svg/
    │   ├── SKILL.md
    │   ├── CLAUDE.md         # Module map
    │   ├── agents/           # Codex metadata
    │   └── scripts/          # TypeScript converter and tests
    └── daftai-gpt-image-2-5-prompt-optimizer/
        ├── SKILL.md
        ├── CLAUDE.md         # Module map
        ├── agents/           # Codex metadata
        ├── scripts/          # Lightweight TypeScript fetcher and tests
        └── references/       # Source rules, local official reference layout, and acceptance cases

Contributing

Feel free to submit issues and pull requests.

License

MIT License

About

No description or website provided.

Topics

Resources

Stars

18 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages