Reconcile stale running scrapes and allow per-scrape delete

docs/media-scraper.md

@@ -40,8 +40,11 @@ Rules:
 - One running scrape at a time. Further requests queue.
 - `stop` marks the job and lets the current page finish.
 - `resume` continues a paused scrape from its stored cursor, so an interrupted walk does not fetch
-  the pages it already has again. The History popup has a **Resume** button and the API a
-  `POST /api/media/scrape/resume` route.
+  the pages it already has again. The History popup has **Resume** and **Delete** buttons and the
+  API a `POST /api/media/scrape/resume` route.
+- On daemon start, scrapes still marked `running` or `queued` on disk are stale (their job died
+  with the previous daemon). They are reconciled to `paused` so the UI tells the truth and Resume
+  can continue them.
 - Every page is an outbound read. Sends from commands keep their own pacing.
 
 ## Pacing and backoff

docs/webui.md

@@ -140,7 +140,8 @@ box live in Chat, not Call, and the connection stats are not chips yet.
   **New scrape…** opens a popup with **Server** and **Group chat**
   tabs: servers expand into their text channels, each channel (and each group chat) has
   **Links** and **Download** buttons, and group chats always use the puppet. **History** lists
-  each scrape with start time, duration, pages, messages, media counts, bytes, files, and previews.
+  each scrape with start time, duration, pages, messages, media counts, bytes, files, and previews,
+  and offers **Resume** for paused ones and **Delete** for anything not running.
   The scrape strip shows live progress with a pulse and **Stop**. Images and GIFs render from the
   cached preview when one exists, then the local file, then the CDN URL; videos and audio are link
   cards with the poster and posted time in the tooltip.