# Content
**Category:** π» Developer Tools
**Repository:** https://github.com/LatentNoise/content
**Views:** 0
**Installs:** 0
**Upvotes:** 0
**Directory Page:** https://allmcps.com/mcp/content
## Description
Turn URLs, files and text into media, transcripts, summaries, translations and PDF.
## Claude Desktop Quick Installation
Heuristic fallback β verify the package name and runner against the repository README before running it. Uses `npx` (confidence: low):
```json
"mcpServers": {
"content": {
"command": "npx",
"args": ["-y","content"]
}
}
```
## Documentation & README
# Content
## Turn URLs, files, and text into media, knowledge, and documents.
**One self-hosted engine. HomeTube client for YouTube β Content Studio for everything else β HomeTube browser extension β MCP server β CLI β Typed SDK β REST API underneath them all**
[](https://github.com/LatentNoise/content/releases/latest)
[](https://github.com/LatentNoise/content/actions/workflows/ci.yml)
[](docs/operations/deployment.md)
[](LICENSE)
[](https://scorecard.dev/viewer/?uri=github.com/LatentNoise/content)
[Quick start](#quick-start) Β·
[HomeTube](#hometube--youtube-focused) Β·
[Coming from HomeTube?](#coming-from-standalone-hometube) Β·
[All the clients](#the-clients) Β·
[Connect an agent](#mcp-server--for-agents-and-ides) Β·
[Read the docs](https://github.com/LatentNoise/content/blob/HEAD/docs/README.md)
Content is a self-hosted engine that turns supported sources into media,
transcripts, summaries, translations, images, metadata, Markdown, PDF, and
more. The **backend** analyzes each source, determines what it can actually
produce, plans and runs the job, records its history, and delivers the results.
Everything else is a **client of one public contract**: **HomeTube** for the
focused YouTube experience, **Content Studio** for general workflows, a
**Chromium extension** for the tab you are already on, the **`content-mcp`**
server for agents, a **CLI** for terminals and cron, a **typed Python SDK** for
applications, the **REST API** for every other language, and **Content
Console** to watch the engine work. Pick one or several β none of them is a
layer the others have to pass through, and none of them holds business logic of
its own.
**HomeTube** is the quickest way to see what that means: paste a YouTube URL,
pick what you want out of it, and watch the files land in your library.
HomeTube in Content β paste a YouTube URL, choose what you want,
follow the job into your library. More about HomeTube β
> [!NOTE]
> **A proven workflow, now growing into a platform.** The standalone
> [HomeTube](https://github.com/EgalitarianMonkey/hometube) container has passed
> **300,000 package downloads on GitHub Container Registry**. It keeps running
> and stays maintained β nothing breaks, and there is no deadline. Content is
> where that work continues, and where its users are invited to move at their
> own pace.
>
> The two are separate projects: standalone HomeTube has not been retrofitted to
> run on Content. The **HomeTube app in this repository is a new UI on the
> Content engine**, carrying the same workflow forward.
> [Coming from standalone HomeTube? β](#coming-from-standalone-hometube)
## Quick start
Docker Compose is the only runtime prerequisite. Install from the published
images β nothing to clone or build:
```bash
mkdir content && cd content
curl -fsSLO https://raw.githubusercontent.com/LatentNoise/content/main/deploy/docker-compose.yml
curl -fsSL -o .env https://raw.githubusercontent.com/LatentNoise/content/main/.env.example
docker compose up -d
```
The default `.env` starts the complete local stack:
| Service | Default URL | Purpose |
| --- | --- | --- |
| **Content Studio** | | General requests across URL, file, upload, and inline-text sources |
| **HomeTube** | | Focused YouTube video and playlist workflow |
| **Content Console** | | Health, configuration, storage, jobs, events, and logs |
| **Content API** | | REST API, OpenAPI, and embedded worker |
Open **Content Studio** for the general workflow, or **HomeTube** if your
source is YouTube. You can also skip both web apps entirely and drive the same
engine from the [MCP server, CLI, SDK, extension, or REST API](#the-clients).
Everything stays in the installation folder: `./data` holds the database, jobs,
and artifacts; finished files are delivered to `./playground/output`. Set
`CONTENT_DELIVERY_DIR_HOST` in `.env` to point at your own NAS or media library
instead β its sub-folders then become the destination choices offered in the
clients.
Update later with:
```bash
docker compose pull && docker compose up -d
```
Build from source instead β for development or to include the
optional speech-to-text runner
```bash
git clone https://github.com/LatentNoise/content.git
cd content
cp .env.example .env
docker compose up -d --build
```
The repository's `docker-compose.yml` adds local `build:` definitions beside
the same images. [`deploy/docker-compose.yml`](https://github.com/LatentNoise/content/blob/HEAD/deploy/docker-compose.yml) is
the build-free deployment file used above; a test keeps the two aligned.
## One source, many artifacts
The same engine serves media and document workflows:
```text
ββ My Conference.mp4
ββ My Conference - audio.opus
YouTube URL βββ Content ββββββΌβ My Conference - subtitles - en.srt
ββ My Conference - subtitles - fr.srt
ββ My Conference - transcript.txt
ββ My Conference - summary.md
ββ My Conference - summary.pdf
Web page / text / .md file βββ Content βββ¬β Article.txt
ββ Article - summary.md
ββ Article - translation.md
ββ Article.pdf
```
Each branch starts with a declarative request. You describe the results;
Content resolves what is possible, plans the work, runs the available tools,
and records where every artifact came from. No yt-dlp flags, ffmpeg pipelines,
transcription glue, or LLM orchestration in the client.
## The clients
One engine, one contract, several independent front doors. Every client below
speaks the same `GenerationRequest`, sees the same resolved capabilities, and
produces the same artifacts β they differ in ergonomics, not in what they can
ask for. HomeTube is entirely optional; so is every other row.
| Client | Best for | Start here |
| --- | --- | --- |
| **[HomeTube](#hometube--youtube-focused)** | The simplest YouTube video or playlist β library experience | Included in Docker Compose at `:8501` |
| **[Content Studio](#content-studio--the-whole-contract-in-a-form)** | General capability-driven requests from URLs, uploads, server files, and inline text | Included in Docker Compose at `:8502` |
| **[Chromium extension](#chromium-extension--the-tab-you-are-already-on)** | Sending the current Chrome, Brave, Edge, or Vivaldi tab to Content | Download the release zip, unzip, *Load unpacked* |
| **[MCP server](#mcp-server--for-agents-and-ides)** | Giving Claude, an IDE, or another MCP agent real artifacts instead of talk | `uv tool install content-mcp` |
| **[CLI](#cli--terminals-scripts-and-cron)** | Terminals, scripts, cron jobs, and raw request files | `uv tool install content-cli` |
| **[Python SDK](#python-sdk--typed-sync-and-async)** | Typed sync and async application integration | `pip install content-sdk` |
| **[REST API](#rest-api--every-other-language)** | Any language or integration that speaks HTTP | `/api/v1`, Swagger at `:8010/docs` |
| **[Content Console](#content-console--observe-and-pilot-the-engine)** | Observing and operating the engine | Included in Docker Compose at `:8503` |
### HomeTube β YouTube, focused
Β· [README](https://github.com/LatentNoise/content/blob/HEAD/apps/web-hometube/README.md)
Paste a URL, choose the media and related artifacts, and follow the job into
your library β the workflow demonstrated at the top of this page.
With HomeTube in Content, you can:
- download a video or playlist as video or audio, with quality, codec,
container, audio-language, and subtitle choices;
- remove or mark sponsored segments with SponsorBlock, cut clips, use
server-side cookie credentials, and embed subtitles, chapters, thumbnails,
and metadata;
- give every artifact a readable name and deliver it into a filesystem library
watched by Plex, Jellyfin, or Emby;
- ask the same source for a transcript, summary, thumbnail, or metadata β not
just the downloaded media β when the required runners are available.
HomeTube is deliberately focused: it does not accept file, upload, or
inline-text sources, and it does not expose the broader document workflow or
PDF output. For those, use Content Studio, the API, CLI, or SDK. **HomeTube is
a Content client, not a layer every Content user must pass through.**
**It has no settings of its own.** It reads the engine's configuration, so what
changes HomeTube lives in the `.env` beside your `docker-compose.yml`: the
delivery library whose sub-folders become the destination choices
(`CONTENT_DELIVERY_DIR_HOST`), the languages pre-selected for audio and
subtitles (`CONTENT_LANGUAGE_PRIMARY`, `CONTENT_LANGUAGES_SECONDARIES`,
`CONTENT_VO_FIRST`, `CONTENT_LANGUAGE_PRIMARY_INCLUDED_IN_SUBTITLES`), and the
server-side cookie file that unlocks age-restricted or members-only videos
(`CONTENT_CREDENTIALS` β HomeTube only ever shows its id; the file never leaves
the server). Nothing is ever pre-selected that the source does not offer, and
no default is final.
[HomeTube's README](https://github.com/LatentNoise/content/blob/HEAD/apps/web-hometube/README.md#configuration) documents each
variable and works the language rules through a concrete example β a Japanese
talk, a French speaker, and exactly what ends up pre-checked. The
[full variable table](https://github.com/LatentNoise/content/blob/HEAD/docs/operations/deployment.md#configuration-a-root-env-not-versioned)
lists everything the engine accepts.
### Coming from standalone HomeTube?
[HomeTube](https://github.com/EgalitarianMonkey/hometube) β the standalone
Streamlit app β **keeps running and stays maintained**. Nothing breaks, and there
is no deadline. Content is where the work continues, so this is a move you make
when you are ready, not one you are forced into.
**What comes across.** The workflow you know: paste a URL, choose quality, codec,
container, audio languages and subtitles; remove or mark sponsored segments with
SponsorBlock; cut clips; unlock restricted videos with a server-side cookie file;
embed subtitles, chapters, thumbnails and metadata; and deliver readable file
names into a library watched by Plex, Jellyfin, or Emby. Playlists come across
too β each entry is analyzed and planned on its own, named per item, and
delivered as its own artifact.
**What is new.** The same source can also produce a transcript, a summary, a
translation, generated thumbnails, metadata, Markdown, or PDF. And the engine is
no longer reachable only from a web form: Studio, the Chromium extension, an MCP
agent, the CLI, the SDK, and plain HTTP all ask for the same things.
**What has not made the trip yet.** HomeTube's playlist **synchronization** β the
plan/apply diff that keeps a local folder in step with a playlist over time
(rename detection, archiving or deleting removed videos, relocation after a
naming or location change) β does not exist here. Content downloads a playlist;
it does not yet re-sync one. It is [roadmapped as part of
M2](docs/roadmap/roadmap.md). If playlist sync is your main use, keep standalone
HomeTube for that job.
**One practical note.** Both HomeTube apps default to port **8501**. Change one of
them in your `.env` if you want to run the two side by side.
### Content Studio β the whole contract, in a form
Β· [README](https://github.com/LatentNoise/content/blob/HEAD/apps/web-studio/README.md)
The general-purpose web app: several sources in one request, every output type
the engine can resolve, with their options, preferences, and constraints. The
form is **capability-driven** β it renders what `/capabilities` answers for
your source, so a capability the engine gains appears without a UI release, and
what cannot be produced is shown with the server's reason instead of silently
disappearing.
Content Studio β build general URL, file, upload and text requests
Studio takes a file two ways, and the distinction matters when the engine runs
on another machine:
- **From this device** β the bytes are sent to the engine and become an
`upload` source ([ADR 0020](https://github.com/LatentNoise/content/blob/HEAD/docs/architecture-decisions/0020-client-uploaded-sources.md)).
Several files at once become several sources. Streamlit buffers the upload in
the UI container, so the ceiling here is deliberately lower than the API's
(200 MB by default, and the picker states the current value rather than
letting you discover it by failing).
- **On the server** β a path the *engine* can already read, under a configured
allowed input root. With this Compose setup that root is `./playground/input`.
For genuinely large files, skip the browser: the SDK streams from disk with
`client.upload_file(path)`.
### Chromium extension β the tab you are already on
[README](https://github.com/LatentNoise/content/blob/HEAD/apps/browser-extension-chromium/README.md) Β·
[latest release](https://github.com/LatentNoise/content/releases/latest)
Send the page you are watching to your engine without opening a UI. Manifest
V3, Chromium only (Chrome, Brave, Edge, Vivaldi, Opera, Arcβ¦), no build step β
what the browser loads is exactly the files in the zip:
1. download `content-browser-extension-chromium-v.zip` from the latest
release and **unzip it** (Chromium loads a folder, not a zip);
2. open `chrome://extensions`, turn on **Developer mode**, **Load unpacked**,
select the folder;
3. open a video and click the extension.
It normalizes the tab's URL (`youtu.be`, Shorts, and embeds become the
canonical watch URL; a `list=` on a watch URL means *that video*, not the
playlist), asks the engine what the source can produce and offers only that,
prefills the name with the naming engine's own proposal, offers the library's
existing folders plus *new folderβ¦*, then follows the job and shows where each
file landed. Every network call happens in the service worker β the engine
sends no CORS headers by default, and that is what lets the extension work
against a stock instance with nothing to configure
([ADR 0016](https://github.com/LatentNoise/content/blob/HEAD/docs/architecture-decisions/0016-first-non-python-client.md)).
Browser extension β send the current tab to Content, from the tab itself
The popup's footer always names the engine it is talking to, so "where is this
sending my video?" is answered on screen. Its README also carries an honest
path-by-path **verification table** β what has actually been driven in a
browser, and what has not.
### MCP server β for agents and IDEs
[README](https://github.com/LatentNoise/content/blob/HEAD/apps/mcp/README.md) Β·
[`content-mcp` on PyPI](https://pypi.org/project/content-mcp/)
Content gives an MCP-compatible agent a controlled way to turn URLs into real
artifacts, not just talk about them. The official **`content-mcp`** server
exposes intention-level tools β not one per endpoint β to:
- analyze a URL and discover what it can produce;
- request one or several outputs and choose a library destination;
- monitor the job, cancel it, and report actionable failures;
- find the artifacts and their delivered paths;
- read small text artifacts inline while keeping large media out of the model
context.
The agent never needs shell access, yt-dlp syntax, or backend internals.
**Nothing to install.** `uvx` fetches the server on first use and caches it, so
the client owns the whole lifecycle (updating stays explicit β `uvx --refresh`,
or pin `content-mcp@x.y.z`):
```bash
# Claude Code
claude mcp add content \
--env CONTENT_API_URL=http://localhost:8010 \
-- uvx content-mcp
```
For Claude Desktop, Cursor, and other MCP clients using the standard JSON
shape:
```json
{
"mcpServers": {
"content": {
"command": "uvx",
"args": ["content-mcp"],
"env": { "CONTENT_API_URL": "http://localhost:8010" }
}
}
}
```
Prefer a pinned executable on your PATH β for an offline machine, or to control
when the version changes? Install it and name it directly
(`"command": "content-mcp"`, or `-- content-mcp` for Claude Code):
```bash
uv tool install content-mcp # or: pipx install content-mcp
```
Then ask naturally:
> Analyze this lecture, save its audio into `Talks`, create a transcript and a
> structured summary if the available runners allow it, and tell me where
> every artifact landed.
The normal tool journey is
`get_config β analyze_source β generate β get_job β get_artifact`. Downloads to
the agent's own machine are bounded by a single variable
(`CONTENT_MCP_DOWNLOAD_DIR`, `~/Downloads/Content` by default): a destination
pointing outside it is refused, not clamped β widening that is the operator's
decision, not something a prompt can talk the server into.
**What it can and cannot do, briefly** β the
[MCP guide](https://github.com/LatentNoise/content/blob/HEAD/apps/mcp/README.md#what-it-supports-today) has the full three
tables, and each row there was driven over stdio against a running engine
rather than inferred:
| | |
| --- | --- |
| **Works** | URLs and playlists Β· local files (uploaded to the engine, so a laptop can drive a NAS) Β· PDFs, read for their text layer Β· video, audio, subtitles, transcripts, summaries, translations, chapters, thumbnails, metadata, Markdown, PDF Β· delivery into your library Β· cookie-authenticated sources |
| **Does not** | no live progress (`get_job` is a poll) Β· no `.docx`/`.epub`/`.odt`/`.rtf`, no OCR for scans Β· no transcoding Β· no playlist sync Β· nothing deletes anything Β· [the API has no authentication](https://github.com/LatentNoise/content/blob/HEAD/docs/architecture-decisions/0024-no-authentication-is-still-the-answer.md) |
| **Needs a runner** | Summaries, translations and derived chapters want a local [Ollama](https://ollama.com) or a cloud key; transcripts want subtitles, or the optional Whisper runner. The engine reports these as `unavailable` up front rather than failing halfway |
| **By design** | **stdio, not an HTTP endpoint** β the server runs where *you* do, which is what lets an agent hand it `~/Documents/report.pdf` and have the bytes uploaded to an engine on your NAS. It also means no open port on a system that has [no authentication](https://github.com/LatentNoise/content/blob/HEAD/docs/architecture-decisions/0024-no-authentication-is-still-the-answer.md). For a client that cannot spawn a process β Open WebUI, hosted UIs β [`mcpo`](https://github.com/open-webui/mcpo) bridges it today |
| **Coming** | [Retrying only what failed](https://github.com/LatentNoise/content/blob/HEAD/docs/architecture-decisions/0025-retrying-only-what-failed.md) Β· [playlist sync](https://github.com/LatentNoise/content/blob/HEAD/docs/roadmap/roadmap.md) Β· [retention](https://github.com/LatentNoise/content/blob/HEAD/docs/architecture-decisions/0023-retention-and-reclaiming-disk.md) Β· more document readers |
### CLI β terminals, scripts, and cron
[README](https://github.com/LatentNoise/content/blob/HEAD/apps/cli/README.md) Β·
[`content-cli` on PyPI](https://pypi.org/project/content-cli/)
```bash
uv tool install content-cli
export CONTENT_API_URL=http://nas.local:8010
content analyze "https://www.youtube.com/watch?v=β¦"
content video "https://β¦" --height 1080 --subs en,fr --watch
content audio "https://β¦" --format opus --playlist --watch
content submit request.json --watch # a raw GenerationRequest, or -
content jobs ; content job ; content artifacts
content download -o out.mkv
```
`--watch` follows the job's **event stream** rather than re-asking for its
status, and the exit code carries the outcome so a script can chain on it
([ADR 0021](https://github.com/LatentNoise/content/blob/HEAD/docs/architecture-decisions/0021-a-job-does-not-succeed-with-a-failed-step.md)):
`0` succeeded, `2` partially succeeded, `1` failed or cancelled. `2` is
deliberately distinct β a playlist that yielded five videos of six is not a
failure, but a script that treats it as a success will quietly move on with
missing files.
### Python SDK β typed, sync and async
[README](https://github.com/LatentNoise/content/blob/HEAD/packages/python-sdk/README.md) Β·
[`content-sdk` on PyPI](https://pypi.org/project/content-sdk/)
The **one** official API client: the CLI, the MCP server, and the web apps all
speak through it, so the engine's rules are never duplicated. Python 3.11+,
`httpx` and `pydantic` only.
```python
from content_sdk import ContentClient, outputs
with ContentClient("http://localhost:8010") as client:
analysis = client.analyze(outputs.url_source("https://www.youtube.com/watch?v=β¦"))
job = client.generate(analysis.id, [outputs.audio_output()])
job.wait()
for artifact in job.artifacts:
print(artifact.display_filename, artifact.delivered_path)
```
`AsyncContentClient` mirrors the whole surface. Analyses are addressable, so
every call accepts either an `analysis_id` or inline sources
([ADR 0014](https://github.com/LatentNoise/content/blob/HEAD/docs/architecture-decisions/0014-addressable-analyses.md));
`client.upload_file(path)` streams a local file to the engine and hands back a
ready-to-use source; and every non-2xx becomes a typed exception carrying the
stable error codes (`NotFound`, `Gone`, `ValidationError`).
### REST API β every other language
[Public contract](https://github.com/LatentNoise/content/blob/HEAD/docs/contract.md) Β· Swagger at
`/api/v1` is the contract every client above is built on, and it is stable
enough to build a ninth client against: requests describe desired outputs
rather than technical operations, error codes are machine identifiers that do
not change (only their human messages do), and *"valid but not implemented"* is
a different, explicit answer from *"invalid"*. Reserved fields are refused
rather than silently ignored. The V1 API has **no built-in authentication** β
see [deployment and security](#deployment-and-security).
### Content Console β observe and pilot the engine
Β· [README](https://github.com/LatentNoise/content/blob/HEAD/apps/web-admin/README.md)
The operations console: strictly observability and control, and deliberately
**not** a way to create downloads (that is Studio and HomeTube).
Content Console β observe and control the engine
