> ## Documentation Index
> Fetch the complete documentation index at: https://insight-flow.potatodog.cc/llms.txt
> Use this file to discover all available pages before exploring further.

# Sources

> The source model, supported collection strategies, and usage flow

Sources form the input layer of Insight Flow. A source defines where data comes from, how it is collected, and which category it belongs to.

## Supported Collection Strategies

* `rss`: public RSS or Atom feeds
* `blog_scraper`: profile-driven parsing for sites without RSS
* `deepbrowse`: fallback for dynamic or complex sites
* `github_trending`: GitHub Trending aggregation
* `huggingface`: Hugging Face Daily Papers aggregation

## Key Capabilities

* Manage preset and custom sources in one place
* Test a single source before using it in a monitor
* Organize sources by category
* Sync preset YAML definitions into the database

## How to Use It

1. Open the Sources page
2. Create or edit a source with name, URL, category, and strategy
3. Run the source test
4. Select that source when creating a monitor

## Related API

* List and detail: `GET /api/v1/sources`, `GET /api/v1/sources/{source_id}`
* Create and update: `POST /api/v1/sources`, `PATCH /api/v1/sources/{source_id}`
* Test: `POST /api/v1/sources/{source_id}/test`
* Delete: `DELETE /api/v1/sources/{source_id}`

## Operational Notes

* Sync presets with `make sync-sources`
* Site profiles live under `backend/app/collectors/site_profiles/`
* For non-RSS sites, prefer the profile-driven flow described in [Collector Plugins](/en/development/collector-plugin)