Add semantic cross-source dedup via local embeddings

- LocalModelClient.embed() calls the OpenAI-compatible /embeddings endpoint
  (local nomic model); base_url shared with chat, model via GOODNEWS_EMBED_MODEL.
- New article_embeddings table and articles.duplicate_of column (+ migration).
- dedup module: embeds missing articles, clusters near-identical stories within
  a date window by cosine similarity (pure-stdlib, vectors normalised once), and
  marks all but the highest-ranked member of each cluster as a duplicate.
- 'dedup' CLI command; cycle now runs poll -> classify -> dedup -> brief.
- Feed and brief queries hide duplicates, so a story carried by multiple
  outlets shows once.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

goodnews/db.py

@@ -37,6 +37,7 @@ CREATE TABLE IF NOT EXISTS articles (
     raw_guid TEXT,
     url_hash TEXT NOT NULL UNIQUE,
     title_hash TEXT,
+    duplicate_of INTEGER REFERENCES articles(id) ON DELETE SET NULL,
     FOREIGN KEY (source_id) REFERENCES sources(id)
 );
 
@@ -62,6 +63,14 @@ CREATE TABLE IF NOT EXISTS article_scores (
     scored_at TEXT NOT NULL DEFAULT CURRENT_TIMESTAMP
 );
 
+CREATE TABLE IF NOT EXISTS article_embeddings (
+    article_id INTEGER PRIMARY KEY REFERENCES articles(id) ON DELETE CASCADE,
+    vector BLOB NOT NULL,
+    dim INTEGER NOT NULL,
+    model TEXT NOT NULL,
+    created_at TEXT NOT NULL DEFAULT CURRENT_TIMESTAMP
+);
+
 CREATE TABLE IF NOT EXISTS ingest_runs (
     id INTEGER PRIMARY KEY AUTOINCREMENT,
     source_id INTEGER REFERENCES sources(id) ON DELETE SET NULL,
@@ -114,7 +123,15 @@ def _migrate(conn: sqlite3.Connection) -> None:
     CREATE TABLE IF NOT EXISTS never alters an existing table, so new columns
     need an explicit, idempotent ALTER guarded by the current column set.
     """
-    cols = {row["name"] for row in conn.execute("PRAGMA table_info(article_scores)")}
+    score_cols = {row["name"] for row in conn.execute("PRAGMA table_info(article_scores)")}
     for column in ("topic", "flavor"):
-        if column not in cols:
+        if column not in score_cols:
             conn.execute(f"ALTER TABLE article_scores ADD COLUMN {column} TEXT")
+
+    article_cols = {row["name"] for row in conn.execute("PRAGMA table_info(articles)")}
+    if "duplicate_of" not in article_cols:
+        conn.execute(
+            "ALTER TABLE articles ADD COLUMN duplicate_of INTEGER REFERENCES articles(id)"
+        )
+    # Created here (not in SCHEMA) so it runs after the column exists on upgrades.
+    conn.execute("CREATE INDEX IF NOT EXISTS idx_articles_duplicate_of ON articles(duplicate_of)")