add re-extract from Strava originals endpoint and improve diag

- POST /api/admin/users/{handle}/reextract-originals: reads stored
  originals/strava/*.json and re-runs strava_to_parsed + ingest_parsed
  without hitting the Strava API; streams SSE progress; calls merge_all
  and rebuild on completion
- GET /api/admin/users/{handle}/diag: now shows _merged/activities/
  file counts, a sample of filenames in activities/ (with symlink flag),
  and lists pending_files by name
- Admin page: Re-extract button per user with live SSE progress modal

bincio/serve/server.py

@@ -619,6 +619,82 @@ async def admin_rebuild_sync(
     return JSONResponse(resp)
 
 
+@app.post("/api/admin/users/{handle}/reextract-originals")
+async def admin_reextract_originals(
+    handle: str,
+    bincio_session: Optional[str] = Cookie(default=None),
+) -> StreamingResponse:
+    """Re-extract activities from stored Strava originals without hitting the API.
+
+    Reads each file in originals/strava/{id}.json (containing {"meta", "streams"}),
+    calls strava_to_parsed, and writes the activity JSON + GeoJSON.
+
+    Skips activities that already have a valid JSON file in activities/.
+    Streams SSE progress; call /rebuild after completion if no --webroot.
+
+    SSE events:
+      {"type": "progress", "n": N, "total": T, "name": "...", "status": "imported"|"skipped"|"error", "detail": "..."}
+      {"type": "done", "imported": N, "skipped": N, "errors": N}
+    """
+    _require_admin(bincio_session)
+    user_dir = _get_data_dir() / handle
+    originals_dir = user_dir / "originals" / "strava"
+    if not originals_dir.exists():
+        raise HTTPException(404, f"No Strava originals directory for '{handle}'")
+
+    original_files = sorted(originals_dir.glob("*.json"))
+    total = len(original_files)
+    log.info("reextract[%s]: starting, %d originals found", handle, total)
+
+    def event_stream():
+        from bincio.extract.strava_api import strava_to_parsed
+        from bincio.extract.ingest import ingest_parsed
+        from bincio.render.merge import merge_all
+
+        imported = 0
+        skipped = 0
+        errors = 0
+        any_imported = False
+
+        for n, orig_path in enumerate(original_files, 1):
+            try:
+                raw = json.loads(orig_path.read_text(encoding="utf-8"))
+                meta = raw.get("meta", {})
+                streams = raw.get("streams", {})
+                name = meta.get("name", orig_path.stem)
+
+                parsed = strava_to_parsed(meta, streams)
+
+                from bincio.extract.writer import make_activity_id
+                activity_id = make_activity_id(parsed)
+                if (user_dir / "activities" / f"{activity_id}.json").exists():
+                    skipped += 1
+                    yield f"data: {json.dumps({'type': 'progress', 'n': n, 'total': total, 'name': name, 'status': 'skipped'})}\n\n"
+                    continue
+
+                ingest_parsed(parsed, user_dir, privacy="public", rdp_epsilon=0.0001)
+                imported += 1
+                any_imported = True
+                yield f"data: {json.dumps({'type': 'progress', 'n': n, 'total': total, 'name': name, 'status': 'imported'})}\n\n"
+            except Exception as exc:
+                errors += 1
+                log.error("reextract[%s]: failed on %s: %s", handle, orig_path.name, exc, exc_info=True)
+                yield f"data: {json.dumps({'type': 'progress', 'n': n, 'total': total, 'name': orig_path.stem, 'status': 'error', 'detail': str(exc)})}\n\n"
+
+        log.info("reextract[%s]: done — imported=%d skipped=%d errors=%d", handle, imported, skipped, errors)
+        if any_imported:
+            merge_all(user_dir)
+            _trigger_rebuild(handle)
+
+        yield f"data: {json.dumps({'type': 'done', 'imported': imported, 'skipped': skipped, 'errors': errors})}\n\n"
+
+    return StreamingResponse(
+        event_stream(),
+        media_type="text/event-stream",
+        headers={"Cache-Control": "no-cache", "X-Accel-Buffering": "no"},
+    )
+
+
 @app.get("/api/admin/users/{handle}/diag")
 async def admin_diag(
     handle: str,
@@ -662,6 +738,25 @@ async def admin_diag(
         except Exception:
             root_activity_count = -1
 
+    # Peek at a few filenames in activities/ to understand the actual state
+    acts_sample: list[str] = []
+    acts_symlinks = 0
+    if activities_dir.exists():
+        for f in sorted(activities_dir.iterdir())[:10]:
+            acts_sample.append(f.name + (" → symlink" if f.is_symlink() else ""))
+            if f.is_symlink():
+                acts_symlinks += 1
+
+    # Check _merged/activities/ separately
+    merged_acts_dir = merged_dir / "activities"
+    merged_acts_json = _count(merged_acts_dir, "*.json")
+    merged_acts_geojson = _count(merged_acts_dir, "*.geojson")
+
+    # List pending files
+    pending_files: list[str] = []
+    if uploads_dir.exists():
+        pending_files = [f.name for f in uploads_dir.iterdir() if f.is_file()]
+
     return JSONResponse({
         "handle": handle,
         "user_dir": str(user_dir),
@@ -669,6 +764,8 @@ async def admin_diag(
             "json_files": _count(activities_dir, "*.json"),
             "geojson_files": _count(activities_dir, "*.geojson"),
             "size_mb": round(_size_mb(activities_dir), 2),
+            "sample": acts_sample,
+            "symlink_count": acts_symlinks,
         },
         "originals": {
             "exists": originals_dir.exists(),
@@ -679,12 +776,15 @@ async def admin_diag(
             "exists": merged_dir.exists(),
             "activity_count_in_index": merged_activity_count,
             "size_mb": round(_size_mb(merged_dir), 2),
+            "activities_json": merged_acts_json,
+            "activities_geojson": merged_acts_geojson,
         },
         "root_index": {
             "exists": root_index.exists(),
             "activity_count": root_activity_count,
         },
-        "pending_uploads": _count(uploads_dir),
+        "pending_uploads": len(pending_files),
+        "pending_files": pending_files,
         "dedup_cache_exists": (user_dir / ".bincio_cache.json").exists(),
         "athlete_json_exists": (user_dir / "athlete.json").exists(),
     })

site/src/pages/admin/index.astro

@@ -31,6 +31,16 @@ import Base from '../../layouts/Base.astro';
       </table>
     </div>
 
+    <!-- Re-extract progress modal -->
+    <dialog id="reextract-dialog" class="rounded-xl bg-zinc-900 border border-zinc-700 p-6 text-white max-w-2xl w-full backdrop:bg-black/60">
+      <div class="flex items-center justify-between mb-4">
+        <h3 class="font-semibold text-sm">Re-extract from Strava originals — <span id="reextract-handle" class="text-zinc-400 font-mono"></span></h3>
+        <button id="reextract-close" class="text-zinc-500 hover:text-zinc-200 text-xs px-2 py-1 rounded bg-zinc-800" disabled>Close</button>
+      </div>
+      <div id="reextract-summary" class="text-xs text-zinc-400 mb-2"></div>
+      <div class="bg-zinc-950 rounded p-3 h-64 overflow-y-auto" id="reextract-log"></div>
+    </dialog>
+
     <!-- Diag modal -->
     <dialog id="diag-dialog" class="rounded-xl bg-zinc-900 border border-zinc-700 p-6 text-white max-w-2xl w-full backdrop:bg-black/60">
       <div class="flex items-center justify-between mb-4">
@@ -62,6 +72,13 @@ import Base from '../../layouts/Base.astro';
   const diagOutput  = document.getElementById('diag-output')!;
   document.getElementById('diag-close')!.addEventListener('click', () => diagDialog.close());
   diagDialog.addEventListener('click', e => { if (e.target === diagDialog) diagDialog.close(); });
+
+  const reextractDialog  = document.getElementById('reextract-dialog') as HTMLDialogElement;
+  const reextractHandle  = document.getElementById('reextract-handle')!;
+  const reextractSummary = document.getElementById('reextract-summary')!;
+  const reextractLog     = document.getElementById('reextract-log')!;
+  const reextractClose   = document.getElementById('reextract-close') as HTMLButtonElement;
+  reextractClose.addEventListener('click', () => { reextractDialog.close(); load(); });
   const confirmOk     = document.getElementById('confirm-ok')!;
   const confirmCancel = document.getElementById('confirm-cancel')!;
 
@@ -136,6 +153,11 @@ import Base from '../../layouts/Base.astro';
                 data-handle="${u.handle}"
                 title="Show diagnostic snapshot of this user's data directory"
               >Diag</button>
+              <button
+                class="reextract-btn text-xs px-3 py-1.5 rounded-lg bg-zinc-800 hover:bg-amber-900 hover:text-amber-300 text-zinc-400 transition-colors"
+                data-handle="${u.handle}"
+                title="Re-extract activities from stored Strava originals (no API call)"
+              >Re-extract</button>
               <button
                 class="rebuild-btn text-xs px-3 py-1.5 rounded-lg bg-zinc-800 hover:bg-zinc-700 text-zinc-400 hover:text-zinc-200 transition-colors"
                 data-handle="${u.handle}"
@@ -157,6 +179,63 @@ import Base from '../../layouts/Base.astro';
       `;
     }).join('');
 
+    tbodyEl.querySelectorAll<HTMLButtonElement>('.reextract-btn').forEach(btn => {
+      btn.addEventListener('click', async () => {
+        const h = btn.dataset.handle!;
+        reextractHandle.textContent = h;
+        reextractLog.innerHTML = '';
+        reextractSummary.textContent = 'Starting…';
+        reextractClose.disabled = true;
+        reextractDialog.showModal();
+
+        let imported = 0, skipped = 0, errors = 0;
+        try {
+          const es = new EventSource(`/api/admin/users/${h}/reextract-originals`);
+          // EventSource only does GET; use fetch + ReadableStream instead
+          es.close();
+
+          const r = await fetch(`/api/admin/users/${h}/reextract-originals`, {
+            method: 'POST', credentials: 'include',
+          });
+          const reader = r.body!.getReader();
+          const decoder = new TextDecoder();
+          let buf = '';
+          while (true) {
+            const { done, value } = await reader.read();
+            if (done) break;
+            buf += decoder.decode(value, { stream: true });
+            const lines = buf.split('\n\n');
+            buf = lines.pop() ?? '';
+            for (const chunk of lines) {
+              const dataLine = chunk.split('\n').find(l => l.startsWith('data: '));
+              if (!dataLine) continue;
+              const ev = JSON.parse(dataLine.slice(6));
+              if (ev.type === 'progress') {
+                const color = ev.status === 'imported' ? 'text-green-400'
+                            : ev.status === 'error'    ? 'text-red-400'
+                            : 'text-zinc-500';
+                const line = document.createElement('div');
+                line.className = `text-xs font-mono ${color}`;
+                line.textContent = `[${ev.n}/${ev.total}] ${ev.status.padEnd(8)} ${ev.name}${ev.detail ? ' — ' + ev.detail : ''}`;
+                reextractLog.appendChild(line);
+                reextractLog.scrollTop = reextractLog.scrollHeight;
+                if (ev.status === 'imported') imported++;
+                else if (ev.status === 'error') errors++;
+                else skipped++;
+                reextractSummary.textContent = `Processing… ${ev.n}/${ev.total} — imported: ${imported}, skipped: ${skipped}, errors: ${errors}`;
+              } else if (ev.type === 'done') {
+                reextractSummary.textContent = `Done — imported: ${ev.imported}, skipped: ${ev.skipped}, errors: ${ev.errors}`;
+              }
+            }
+          }
+        } catch (err) {
+          reextractSummary.textContent = 'Error: ' + String(err);
+        } finally {
+          reextractClose.disabled = false;
+        }
+      });
+    });
+
     tbodyEl.querySelectorAll<HTMLButtonElement>('.diag-btn').forEach(btn => {
       btn.addEventListener('click', async () => {
         const h = btn.dataset.handle!;