# Sub-goal: Build the local webapp (Workstream B)

**Parent goal:** Convert yt-channel-scraper into a local content-mining platform.
**Owner agent:** `webapp-builder` (`.opencode/agent/webapp-builder.md`)
**Spec:** `docs/superpowers/specs/2026-07-26-platform-design.md` (§7 Webapp, §8 Cookies, §7.2 design language)
**Status:** Workstream A (engine + CLI) is DONE and verified — `pytest` green, real DB backfilled, FTS search working, cookie vault auto-imports the real cookie.

## Objective
Implement the FastAPI backend + Alpine.js SPA so the platform is usable from a browser, launched by `start-server.bat`. No Node/npm build step.

## Deliverables
1. `src/yt_scraper/webapp/__init__.py`
2. `src/yt_scraper/webapp/app.py` — FastAPI factory, `/healthz`, static mount, auto-import cookies on startup.
3. `src/yt_scraper/webapp/jobs.py` — single-worker scrape job runner with SSE event bus (queue/progress/log/cancel), persists to `scrape_jobs`.
4. `src/yt_scraper/webapp/api.py` — REST routers documented in spec §7.1 (dashboard, channels, videos, search, analysis, scrape+SSE, cookies with drag-and-drop upload, exports).
5. `src/yt_scraper/webapp/static/index.html`, `app.js`, `styles.css` — the 9-view SPA, dark command-center theme, drag-and-drop cookie vault, Chart.js viz.

## API contract (the SPA consumes exactly this)
- `GET  /healthz` → `{"status":"ok"}`
- `GET  /api/dashboard` → `{channels[], status_breakdown{}, uploads_over_time[], duration_histogram[], top_tags[]}`
- `GET  /api/channels` | `POST /api/channels {url}` | `DELETE /api/channels/{id}`
- `GET  /api/videos?channel=&status=&from=&to=&min_dur=&q=&sort=&page=&size=` → `{items[], total, page, size}`
- `GET  /api/videos/{id}` | `GET /api/videos/{id}/transcript` → `[{idx,start_sec,end_sec,text}]`
- `GET  /api/search?q=&channel=&limit=` → `[{video_id,title,start_sec,snippet}]`
- `GET  /api/analysis?channel=&type=topwords|wordcloud|timeline&term=` → chart-ready JSON
- `POST /api/scrape {channel_id,opts}` → `{job_id}` | `GET /api/scrape/{id}/stream` (SSE) | `DELETE /api/scrape/{id}`
- `POST /api/cookies/upload` (multipart file(s) or `{text,label}` JSON) | `GET /api/cookies` | `POST /api/cookies/{id}/activate` | `DELETE /api/cookies/{id}` | `POST /api/cookies/{id}/test`
- `GET  /api/export?format=json|csv|srt|html&channel=` → file stream

## Done when
- `/healthz` returns 200; every router smokes green against the real DB; SPA renders all 9 views; SSE streams a scrape job; the real cookie drag-and-drops and activates; `pytest` green.