docs: update v0.19.0 release notes with recent commits

Add write_note overwrite guard (#632) to new capabilities, 8 bug fixes
(#631, #630, #30, #31, #28, plus schema_validate/Post/frontmatter fixes),
and write_note idempotency breaking change to upgrade notes. Bump commit
count from 80+ to 90+.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: phernandez <paul@basicmachines.co>

Author: phernandez

docs/releases/v0.19.0.md

@@ -4,7 +4,7 @@
 
 v0.19.0 is a major release that introduces semantic vector search, a schema validation system,
 project-prefixed permalinks, per-project cloud routing, and a significant upgrade to FastMCP 3.0.
-It includes 80+ commits since v0.18.0 spanning new features, architectural improvements, and
+It includes 90+ commits since v0.18.0 spanning new features, architectural improvements, and
 stability fixes across both SQLite and Postgres backends.
 
 ---
@@ -138,6 +138,13 @@ Search results now surface more relevant context:
 - `TOP_CHUNKS_PER_RESULT` increased from 3 to 5, catching answers deeper in large notes (~2700 → ~4500 chars)
 - `CONTENT_DISPLAY_LIMIT` doubled from 2000 to 4000 chars for results without matched chunks
 
+### `write_note` Overwrite Guard (#632)
+
+`write_note` is now non-idempotent by default. If a note already exists, the tool returns an
+error instead of silently overwriting. Pass `overwrite=True` to replace, or use `edit_note`
+for incremental updates. Config option `write_note_overwrite_default` restores the old upsert
+behavior.
+
 ---
 
 ## Architecture Changes
@@ -242,6 +249,14 @@ making all major CLI commands scriptable for CI pipelines and automation.
 - **#606**: Accept `null` for `expected_replacements` in `edit_note`
 - **#579, #607**: Guard against closed streams in promo panel and missing vector tables on shutdown
 - **#609**: FTS-only hybrid results missing `matched_chunk_text`; content limits too conservative
+- **#631**: `build_context` related_results schema validation failure — replaced fragile `_slim_context()` stripping with Pydantic `exclude=True` field config
+- **#630**: Skip workspace resolution when client factory is active — prevents 401 errors in cloud MCP server mode
+- **#30**: `tag:` prefix query fails with hybrid search — moved tag prefix parsing to MCP tool level so it works with all search modes
+- **#31**: `search_notes` returns cluttered observation/relation-level results — now defaults to entity-level results
+- **#28**: `schema_infer` and `schema_diff` return raw Pydantic models as "undefined" in LLM output — added markdown formatters
+- Fix `schema_validate` identifier resolution (now uses LinkResolver) and text rendering (markdown formatter)
+- Fix `Post(**metadata)` crash when frontmatter contains `content` or `handler` keys
+- Fix list-valued frontmatter fields (`title`, `type`) crashing on `.strip()` — now coerced to strings
 - Cap sqlite-vec knn `k` parameter at 4096 to prevent backend errors
 - Parameterize SQL queries in search repository type filters
 - Double-default display in project list
@@ -297,3 +312,6 @@ making all major CLI commands scriptable for CI pipelines and automation.
   added on next sync. Set `ensure_frontmatter_on_sync: false` to preserve old behavior.
 - **Config migration** runs automatically for cloud projects with bisync — `local_sync_path`
   is promoted to `path` so filesystem operations work correctly.
+- **`write_note` is no longer idempotent** — calls to `write_note` for existing notes now
+  return an error unless `overwrite=True` is passed. Use `edit_note` for incremental changes,
+  or set `write_note_overwrite_default: true` in config to restore the old behavior.