fix(node): skip indexing audio files under 50KB, likely-corrupt source

The "P.H. Theme" failure investigated earlier turned out to be a genuinely
corrupt 1256-byte source file with no audio stream at all, just an ID3
tag — a real, if rare, corruption pattern worth guarding against directly
rather than only handling gracefully at playback time. Scoped to audio
only, applied wherever a file actually gets hashed/typed (fresh scan and
the cache-miss rehash path alike) — a tiny file of any other type is still
indexed normally.

3 files changed, 58 insertions, 2 deletions

diff --git a/packages/meshbay-node/src/meshbay_node/indexer/indexer.py b/packages/meshbay-node/src/meshbay_node/indexer/indexer.py
index 3ca34f6..10981c0 100644
--- a/packages/meshbay-node/src/meshbay_node/indexer/indexer.py
+++ b/packages/meshbay-node/src/meshbay_node/indexer/indexer.py
@@ -79,6 +79,19 @@ def _is_indexable(path: Path) -> bool:
     )
 
 
+# Found live: a 1256-byte ".mp3" with no audio stream at all, just an ID3
+# tag — a truncated/corrupted rip, sitting between two good tracks of the
+# same album (docs/musicbay.md). A source this small claiming to be audio
+# is far more likely broken than real, so it is skipped before ever being
+# hashed rather than indexed and left to fail at playback time. Scoped to
+# audio only — a tiny real file of any other type is still worth indexing.
+MIN_AUDIO_SIZE_BYTES = 50 * 1024
+
+
+def _is_indexable_size(path: Path, size: int) -> bool:
+    return not (_detect_type(path) == "audio" and size < MIN_AUDIO_SIZE_BYTES)
+
+
 _HASH_CHUNK = 8 * 1024 * 1024   # 8 MB streaming hash chunks
 
 
@@ -134,6 +147,8 @@ def _scan_file(root: Root, file_path: Path) -> IndexEntry | None:
         return None
     try:
         stat = file_path.stat()
+        if not _is_indexable_size(file_path, stat.st_size):
+            return None
         hasher = blake3.blake3()
         # long_path is a no-op off Windows; there it is what lets a deep media
         # library past MAX_PATH.
@@ -338,6 +353,8 @@ class DirectoryIndexer:
             st = file_path.stat()
         except OSError:
             return None
+        if not _is_indexable_size(file_path, st.st_size):
+            return None
 
         if self._cache is not None:
             cached = await self._cache.lookup(str(file_path), st.st_size, st.st_mtime)
diff --git a/packages/meshbay-node/tests/test_indexer.py b/packages/meshbay-node/tests/test_indexer.py
index 56571a4..4bae620 100644
--- a/packages/meshbay-node/tests/test_indexer.py
+++ b/packages/meshbay-node/tests/test_indexer.py
@@ -28,7 +28,11 @@ def shared_dir(tmp_path):
     d = tmp_path / "shared"
     d.mkdir()
     (d / "video.mkv").write_bytes(os.urandom(1024))
-    (d / "music.mp3").write_bytes(os.urandom(512))
+    # Above MIN_AUDIO_SIZE_BYTES: these tests use this file as a generic
+    # "here's an audio entry" stand-in, not as a test of the tiny-file gate
+    # itself (see test_is_indexable_size_* / test_initial_scan_skips_a_
+    # corrupt_tiny_audio_file below for that).
+    (d / "music.mp3").write_bytes(os.urandom(indexer_mod.MIN_AUDIO_SIZE_BYTES + 1024))
     (d / "readme.md").write_bytes(b"# Hello MeshBay")
     subdir = d / "docs"
     subdir.mkdir()
@@ -157,6 +161,37 @@ def test_type_detection_covers_wma_and_musepack():
     assert indexer_mod._detect_type(Path("track.mpc")) == "audio"
 
 
+def test_is_indexable_size_rejects_tiny_audio(tmp_path):
+    assert indexer_mod._is_indexable_size(Path("track.mp3"), 1256) is False
+    assert indexer_mod._is_indexable_size(
+        Path("track.mp3"), indexer_mod.MIN_AUDIO_SIZE_BYTES) is True
+    assert indexer_mod._is_indexable_size(
+        Path("track.mp3"), indexer_mod.MIN_AUDIO_SIZE_BYTES - 1) is False
+
+
+def test_is_indexable_size_does_not_apply_to_other_types():
+    """A tiny document/image is still worth indexing — the rule exists
+    because a truncated *audio* file is a known corruption signature, not
+    because small files in general are suspect."""
+    assert indexer_mod._is_indexable_size(Path("readme.txt"), 10) is True
+    assert indexer_mod._is_indexable_size(Path("icon.png"), 10) is True
+
+
+@pytest.mark.asyncio
+async def test_initial_scan_skips_a_corrupt_tiny_audio_file(tmp_path, sk_node, gek):
+    d = tmp_path / "shared"
+    d.mkdir()
+    (d / "good.mp3").write_bytes(os.urandom(indexer_mod.MIN_AUDIO_SIZE_BYTES + 1))
+    (d / "corrupt.mp3").write_bytes(os.urandom(1256))
+
+    indexer = DirectoryIndexer(roots=one_root(d), group_id="g", sk_node=sk_node, gek=gek)
+    await indexer.initial_scan()
+
+    names = {e.name for e in indexer.index.entries}
+    assert "good.mp3" in names
+    assert "corrupt.mp3" not in names
+
+
 @pytest.mark.asyncio
 async def test_hidden_files_excluded(tmp_path, sk_node, gek):
     d = tmp_path / "dir"
diff --git a/packages/meshbay-node/tests/test_root_availability.py b/packages/meshbay-node/tests/test_root_availability.py
index 028c308..0201c1f 100644
--- a/packages/meshbay-node/tests/test_root_availability.py
+++ b/packages/meshbay-node/tests/test_root_availability.py
@@ -14,6 +14,7 @@ straightforward implementation does.
 """
 
 import asyncio
+import os
 from pathlib import Path
 
 import pytest
@@ -90,7 +91,10 @@ async def test_one_root_going_away_leaves_the_others_alone(tmp_path):
     films.mkdir()
     music.mkdir()
     (films / "a.mkv").write_bytes(b"a")
-    (music / "b.mp3").write_bytes(b"b")
+    # Above the tiny-audio-file cutoff (indexer.py's MIN_AUDIO_SIZE_BYTES) —
+    # this test is about root availability, not that gate, so the content
+    # just needs to actually get indexed as an entry.
+    (music / "b.mp3").write_bytes(os.urandom(60 * 1024))
 
     idx = await _indexer(_roots(films, music))
     assert _names(idx) == {"a.mkv", "b.mp3"}