feat: add seo_meta service for AI-generated SEO metadata with CLI management and database layer

Implement a new `SeoMetaService` subservice that generates clean SEO title/description/keywords for published content items, distinct from the existing SEO diagnostics auditor. Add `seo_metadata` polymorphic table with soft-delete support, batch query methods, and usage tracking. Extend the CLI with `seo-meta prune` and `seo-meta clear` commands for job row lifecycle management. Wire `schedule_seo_meta_for_table` into content creation and editing flows in `content.py`. Document the new service in `AGENTS.md` and `README.md`, including the `extra_head` site setting for custom `<head>` injection.
2026-06-19 20:15:22 +00:00
parent 426d3639c6
commit d10f1af118
51 changed files with 2262 additions and 93 deletions
@@ -42,6 +42,7 @@ class VectorStore:
        self.collection_name = collection_name
        self._client = None
        self._collection = None
+        self._dims: int | None = None

    def _ensure(self):
        if self._collection is not None:
@@ -55,9 +56,43 @@ class VectorStore:
        )
        return self._collection

+    @property
+    def dims(self) -> int | None:
+        if self._dims is not None:
+            return self._dims
+        try:
+            collection = self._ensure()
+            data = collection.get(include=["embeddings"], limit=1)
+            rows = data.get("embeddings") or []
+            if rows and rows[0]:
+                self._dims = len(rows[0])
+        except Exception:
+            return self._dims
+        return self._dims
+
    def add(self, chunks: list[Chunk], vectors: list[list[float]]) -> None:
        if not chunks:
            return
+        keep_chunks: list[Chunk] = []
+        keep_vectors: list[list[float]] = []
+        for chunk, vector in zip(chunks, vectors):
+            if not vector:
+                continue
+            if self._dims is None:
+                self._dims = len(vector)
+            if len(vector) != self._dims:
+                logger.warning(
+                    "deepsearch dropping chunk with mismatched embedding dim %d != %d",
+                    len(vector),
+                    self._dims,
+                )
+                continue
+            keep_chunks.append(chunk)
+            keep_vectors.append(vector)
+        if not keep_chunks:
+            return
+        chunks = keep_chunks
+        vectors = keep_vectors
        collection = self._ensure()
        collection.add(
            ids=[chunk.uid for chunk in chunks],