fix: TransferEncodingError on task extraction failure

- brain.py: unwrap JSON array responses from Claude (occasional array
  instead of object caused AttributeError → broken chunked stream)
- main.py: wrap extract_task_fields() in try/except so any extraction
  failure gracefully falls back to regex — generator always completes
- tools.py: _resolve_project() uses alphanumeric-stripped matching so
  "Generaladm" resolves to General/Admin via identifier prefix match.
  Word-overlap fallback added. Default changed from Agent Ecosystem to
  General/Admin (correct catch-all bucket)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

app/brain.py

@@ -69,6 +69,11 @@ async def extract_task_fields(message: str) -> tuple[dict, dict]:
             if content.startswith("json"):
                 content = content[4:]
         fields = json.loads(content.strip())
+        # Claude occasionally wraps the object in an array — unwrap it
+        if isinstance(fields, list):
+            fields = fields[0] if fields else {}
+        if not isinstance(fields, dict):
+            fields = {}
         return fields, usage
     except Exception as e:
         logger.warning(f"extract_task_fields failed: {e}")