subtitle-scout-wizard skill
A skill that walks an agent through deploying the self-hosted Subtitle Scout, from an empty host to the first scan.
- 148 ★ on the repository
- Last commit
- AGPL-3.0
Broad access
Why this level
- The agent runs containers and commands on your server and handles keys for external services.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/subtitle-scout-wizardDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- The agent runs containers and commands on your server and handles keys for external services.
This is third-party code. Review the repository files before installing.
What it does
The skill leads an agent through nine deployment steps for Subtitle Scout: interview, container start, keys, subtitle sources, media folders, a first scan and diagnostics. Everything the agent can do itself through the shell and the container HTTP API it does itself. The user only registers accounts and hands over keys. Subtitle Scout itself finds subtitles for films and episodes with an LLM agent. It matches candidates to the exact episode, checks them in a sandbox, installs the fit and translates when no source has it. The skill includes references for the API, per-language keys and diagnostic trees for startup failures, sources and empty scans.
Who it is for. Owners of home media libraries on a NAS or server who want an agent to deploy Subtitle Scout without learning Docker and the API by hand.
Good fit when
- Subtitle Scout must be set up from zero on a local machine, NAS or remote server.
- Subtitle sources fail validation and the keys look suspect.
- Folder scans return nothing.
- You need to add a media folder or change the target subtitle language.
Not a fit when
- You want a ready-made subtitle service without self-hosting.
- Docker is unavailable on the target host.
- An agent must not run commands on the server holding your media library.
Example request
Use the subtitle-scout-wizard skill to deploy Subtitle Scout on my NAS.Limitations
Needs Docker and network access. Subtitle Scout requires a TMDB key and an LLM provider key; other sources are optional. Matching quality depends on the model tier: weak models err without warning. The SubHD and Zimuku sources are off by default because of terms of service, see `SECURITY.md`. The service license is AGPL-3.0. The skill covers only Subtitle Scout and is no use for other tasks.
How to disable. Remove the skill from the agent directory. Stop the service itself with `docker compose down` in the repository directory; data and cache are deleted separately.
FAQ
What does the agent do and what does the user do?
The agent starts the container, calls the API and validates keys. The user registers accounts, obtains keys and sets the admin password in a browser.
Where do the keys go?
Into the service database through the API, not into `.env`. Only `TZ` and TMDB network settings stay in `.env`; keys in environment variables do nothing.
How do I check that a key works?
Only `POST /api/v2/setup/validate` gives a verdict. Your own requests are fine only to test network reachability, not to judge a key.
How do I install the skill?
Run `npx skills add fancydirty/subtitle-scout`, then tell the agent: use the subtitle-scout-wizard skill to deploy Subtitle Scout.
What does the service need to run?
Docker, a TMDB key and an OpenAI-compatible LLM provider key. Subtitle sources are optional, but configure those for your language.
Discussion
No questions or reviews yet.
Related
SkillsSkill that turns a video into tall subtitle collages: crops burned-in subtitles or draws reviewed lines onto real frames
A skill that transcribes video, lets you review captions in a browser and burns in Chinese and English captions.
A Codex and Claude Code skill: subtitles and a summary from a Douyin, Xiaohongshu, Bilibili or YouTube link or a local file
ChatGPT short-video editing skills
ChatGPT 剪短影音 Skills
Two agent skills: set up the editing environment and build a vertical short video in eight verified steps
Twelve skills teach an agent to read YouTube through the TranscriptOut service: transcripts, search and channels in one HTTPS call
VideoDB skills and an MCP server teach an agent to upload video, cut clips, add subtitles and find moments by meaning
README in short
The service works without Plex, Jellyfin or Sonarr: it scans folders directly. Credentials stay in the app's own SQLite on your machine.