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}")

app/main.py

@@ -686,8 +686,12 @@ async def chat_completions(req: ChatRequest):
 
         elif intent == "task":
             # LLM extraction — handles any natural language phrasing
-            fields, usage = await extract_task_fields(user_message)
-            log_usage("task_extract", usage["prompt_tokens"], usage["completion_tokens"])
+            try:
+                fields, usage = await extract_task_fields(user_message)
+                log_usage("task_extract", usage["prompt_tokens"], usage["completion_tokens"])
+            except Exception as e:
+                logger.warning(f"extraction failed, falling back to regex: {e}")
+                fields = {}
 
             title = fields.get("title") or extract_task_title(user_message)
             project_hint = fields.get("project") or extract_project_name(user_message)

app/tools.py

@@ -147,14 +147,33 @@ async def get_plane_issues(project_hint: str) -> tuple[str, list[dict]]:
 
 
 def _resolve_project(projects: list[dict], hint: str | None) -> dict:
+    import re as _re
     if hint:
         hint_lower = hint.lower()
+        hint_alpha = _re.sub(r"[^a-z0-9]", "", hint_lower)  # "generaladm" from "Generaladm"
+
         for p in projects:
-            if hint_lower in p["name"].lower():
+            name_lower = p["name"].lower()
+            ident_lower = p.get("identifier", "").lower()
+            name_alpha = _re.sub(r"[^a-z0-9]", "", name_lower)  # "generaladmin"
+            # Exact or substring name match
+            if hint_lower in name_lower or name_lower in hint_lower:
                 return p
-    # Default: Agent Ecosystem
+            # Identifier match (e.g. "generaladm" == "generaladm")
+            if hint_alpha == ident_lower or hint_alpha == _re.sub(r"[^a-z0-9]", "", ident_lower):
+                return p
+            # Stripped prefix match (e.g. "generaladm" starts "generaladmin")
+            if name_alpha.startswith(hint_alpha) or hint_alpha.startswith(name_alpha):
+                return p
+        # Word overlap fallback — any significant word of hint in project name
+        hint_words = {w for w in hint_lower.split() if len(w) > 2}
+        for p in projects:
+            if hint_words & set(p["name"].lower().split()):
+                return p
+
+    # Default: General / Admin (catch-all bucket)
     for p in projects:
-        if "agent" in p["name"].lower():
+        if "general" in p["name"].lower():
             return p
     return projects[0]