Replace live spinner with periodic progress prints

console.status doesn't flush to a redirected log, making long background
extract/embed/summarize runs look silent. Print periodic progress lines that
flush instead, so background jobs are observable via their output file.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>

clawlibrary/__main__.py

@@ -154,7 +154,7 @@ def cmd_extract(args: argparse.Namespace) -> int:
         vlm = lambda png: oc.caption_image(png, VLM_OCR_PROMPT)  # noqa: E731
 
     done = skipped = failed = vlm_pages = 0
-    with console.status("Extracting text...") as status:
+    total = len(pdfs)
     for i, pdf in enumerate(pdfs, 1):
         if not args.force and already_extracted(pdf):
             skipped += 1
@@ -166,8 +166,8 @@ def cmd_extract(args: argparse.Namespace) -> int:
         except Exception as e:  # noqa: BLE001
             failed += 1
             console.print(f"[red]✗[/red] {pdf.name}: {e}")
-            if i % 25 == 0:
+        if i % 100 == 0:
-                status.update(f"Extracting... {i}/{len(pdfs)} (done {done}, skipped {skipped})")
+            console.print(f"  extract {i}/{total}  (done {done}, skipped {skipped})")
     console.print(
         f"[green]Extract complete[/green]  extracted={done} skipped={skipped} "
         f"failed={failed} vlm_pages={vlm_pages}"
@@ -193,7 +193,7 @@ def cmd_summarize(args: argparse.Namespace) -> int:
         pdfs = pdfs[: args.limit]
 
     done = skipped = 0
-    with console.status("Summarizing...") as status:
+    total = len(pdfs)
     for i, pdf in enumerate(pdfs, 1):
         side = pdf.with_suffix(".json")
         if not args.force and has_summary(side):
@@ -203,7 +203,8 @@ def cmd_summarize(args: argparse.Namespace) -> int:
         if summary is not None:
             append_insight(insights, _json.loads(side.read_text()))
             done += 1
-            status.update(f"Summarizing... {i}/{len(pdfs)} (done {done})")
+        if i % 10 == 0:
+            console.print(f"  summarize {i}/{total}  (done {done}, skipped {skipped})")
     console.print(f"[green]Summaries complete[/green]  done={done} skipped={skipped}")
     console.print(f"Insight digest: {insights}")
     return 0
@@ -228,7 +229,7 @@ def cmd_embed(args: argparse.Namespace) -> int:
         pdfs = pdfs[: args.limit]
 
     added = skipped = 0
-    with console.status("Embedding...") as status:
+    total = len(pdfs)
     for i, pdf in enumerate(pdfs, 1):
         data = json.loads(pdf.with_suffix(".json").read_text())
         doi = data.get("doi") or data.get("id") or pdf.stem
@@ -245,7 +246,7 @@ def cmd_embed(args: argparse.Namespace) -> int:
         added += 1
         if i % 20 == 0:
             store.save(root)  # periodic checkpoint for resumability
-                status.update(f"Embedding... {i}/{len(pdfs)} (added {added})")
+            console.print(f"  embed {i}/{total}  (added {added}, vectors {store.vectors.shape[0]})")
     store.save(root)
     console.print(
         f"[green]Embedding complete[/green]  papers_added={added} skipped={skipped} "

embed.out

@@ -0,0 +1 @@
+Embedding complete  papers_added=1199 skipped=3 vectors=40978