# media-scrape

The media archive feature. Contained here on purpose: the daemon mounts it, but the logic does not
spread through `poc/src/`.

## Layout

```
media-scrape/
  README.md          this file
  src/               scrape engine, store, extractor, Web UI routes, MCP tools
  slopbot/           the bot core (discord.js backend)
  test/              unit tests against the mock core and a fake bot REST
  db/                runtime data, ignored by git
```

`db/` holds one directory per scrape: `scrape.json`, `index.db`, `media/`, `thumbs/`. It is
configurable with `VESKSLOP_MEDIA_DB_DIR` and never committed. See
[../docs/media.md](../docs/media.md).

The daemon imports this folder. `media-scrape/` is its own pnpm workspace root (like `poc/`), so
`discord.js` is declared in `media-scrape/package.json` and resolves from `media-scrape/node_modules`.

## Docs

- [../docs/media.md](../docs/media.md) — the feature, storage, limits, Web UI
- [../docs/media-scraper.md](../docs/media-scraper.md) — the walk, pacing, schema
- [../docs/slopbot.md](../docs/slopbot.md) — the bot core and its setup
- [../docs/media-mcp.md](../docs/media-mcp.md) — the MCP surface

## Status

Design approved 2026-09-16. Implementation checkpoints live in [../CURRENT.md](../CURRENT.md).