katareayush-skills
Caption videos locally with burned-in subtitles, and generate schema-aware test-data seed scripts for any repo.
Open source Open in the app JSON README (API)
About
Caption videos locally with burned-in subtitles, and generate schema-aware test-data seed scripts for any repo.
Details
- Kind
- Plugins
- Topic
- No topic detected
- Publisher
- katareayush
- Origin
- gemini
- Category
- ferramentas
- Version
- 1.4.0
- Stars
- 8
- Last push
- 2026-07-29T05:34:12Z
- Repository state
- ativo
- Language
- Python
- License
- MIT
- Added
- 2026-08-30 14:13:39
- Updated
- 2026-08-30 14:13:39
- Origin id
katareayush/video-captions
README
# katareayush — agent skills Two skills — **video-captions** and **seedgen** — that run in **every major AI coding agent**, not just one. They're built on [Agent Skills](https://agentskills.dev), the open standard Anthropic published in December 2025 and that Claude Code, Codex CLI, Gemini CLI, GitHub Copilot, Cursor, Windsurf and OpenCode have all adopted. One `SKILL.md`, every host. For agents that only speak MCP, there's a server too. ## Install **Everywhere at once** — detects the agents you have and installs into each: ``` git clone https://github.com/katareayush/video-captions cd video-captions python3 install.py ``` Or without cloning: ``` curl -fsSL https://raw.githubusercontent.com/katareayush/video-captions/main/install.py | python3 ``` Then, in any agent, in any project: ``` add captions to demo.mp4 ``` <details> <summary><b>Per-agent instructions</b></summary> The one-liner above already covers most of these — it installs into every agent it finds. These are the cases that differ. Each is self-contained; none assumes you cloned the repo. | Agent | Install | |-------|---------| | **Claude Code** | `/plugin marketplace add katareayush/video-captions` then `/plugin install video-captions@katareayush` | | **Gemini CLI** | `gemini extensions install https://github.com/katareayush/video-captions` | | **Codex · Copilot · Cursor · OpenCode** | `curl -fsSL $INSTALL \| python3 - --only agents` → `~/.agents/skills`, which all four read | | **Windsurf** | `curl -fsSL $INSTALL \| python3 - --only windsurf` | | **Zed · ChatGPT desktop · n8n · LM Studio** | `curl -fsSL $INSTALL \| python3 - --mcp` prints an MCP config to paste | | **This repo only** | `curl -fsSL $INSTALL \| python3 - --project` → `./.agents/skills` | | **No agent at all** | after setup, `video-captions demo.mp4` works as a plain CLI | where `$INSTALL` is `https://raw.githubusercontent.com/katareayush/video-captions/main/install.py`. Note the `-` after `python3` — it tells Python to read the script from the pipe and pass the rest through as arguments. Other useful flags: `--list` (show what's detected), `--all` (install everywhere regardless), `--uninstall` (remove everything it installed), `--no-deps` (skip ffmpeg/faster-whisper). From a clone, drop the `curl` and just run `python3 install.py --only agents`. </details> ## video-captions Give any agent a video and say *"add captions"* — it transcribes the speech **locally** and burns accurately-timed captions in. The look defaults to a clean, minimal style, and you can **describe any design in plain words** and the agent applies it. Fully offline after a one-time model download, cross-platform, no API key. The first run installs `faster-whisper` + a subtitle-capable `ffmpeg` and downloads the Whisper model once (~150 MB). Everything after is offline. **Requirements:** Python 3.9+, and ffmpeg — setup installs a subtitle-capable build automatically on macOS (`ffmpeg-full`); Windows/Linux ffmpeg already includes it. ### Describe the look | You say | The agent runs | |---------|----------------| | (nothing) | clean white subtitles, bottom-centre | | "big bold yellow captions at the top" | `--pos top --size large --weight bold --color yellow` | | "white text, thick black outline" | `--color white --outline-color black --outline 4` | | "minimal, no shadow" | `--shadow 0` | | "make it TikTok style" | `--word-by-word` | | "3 lowercase words on a rounded pill" | `--preset` | | "2 words at a time on a black plate" | `--words 2 --plate --plate-color black` | | "it's in Hindi, give me English subs" | `--translate` | `--preset` is the one built-in look: editorial lowercase captions, three words at a time on a dark rounded pill, made for light footage. It's just a bundle of individual flags, all of which stay available on their own. Full flag reference: [`plugins/video-captions/README.md`](plugins/video-captions/README.md). ## seedgen Point an agent at any repo and ask it to seed the database — it reads your real schema (SQL migrations, Prisma, Django, Mongoose, Solidity, etc.) and generates a **runnable, schema-aware test-data seed script** that respects enums, NOT NULL, unique, and foreign-key constraints. Stack-agnostic: web2, web3, Python, or anything with a data model. On Claude Code and Gemini CLI it's also a slash command: ``` /seed /seed indian 50 users /seed web3 only ``` A **locale/filter** changes generated values (names, phones, addresses, currency) to match a region without touching the schema. Details in [`plugins/seedgen/README.md`](plugins/seedgen/README.md). ## Repo layout ``` . ├── skills/ # ← source of truth: portable Agent Skills │ ├── video-captions/ │ │ ├── SKILL.md │ │ ├── scripts/ # setup, caption, transcribe, build_captions, │ │ └── requirements.txt # ffmpeg_tools, mcp_server │ └── seedgen/ │ ├── SKILL.md │ └── references/locales.md ├── install.py # installs into every agent on the machine ├── tools/sync.py # regenerates plugins/ from skills/ ├── gemini-extension.json # `gemini extensions install <repo>` ├── commands/ # Gemini slash commands (/captions, /seed) ├── docs/ # the docs site ├── .claude-plugin/ # Claude Code marketplace catalog └── plugins/ # GENERATED — the Claude Code plugin build ``` `skills/` is what you edit. `plugins/` is derived from it by `tools/sync.py` and exists only so existing `/plugin install` users keep working — `python3 tools/sync.py --check` fails if the two drift. ## License MIT — see [LICENSE](LICENSE).