feat: add API provider selection UI and fix stuck features on agent crash

API Provider Selection:
- Add provider switcher in Settings modal (Claude, Kimi, GLM, Ollama, Custom)
- Auth tokens stored locally only (registry.db), never returned by API
- get_effective_sdk_env() builds provider-specific env vars for agent subprocess
- All chat sessions (spec, expand, assistant) use provider settings
- Backward compatible: defaults to Claude, env vars still work as override

Fix Stuck Features:
- Add _cleanup_stale_features() to process_manager.py
- Reset in_progress features when agent stops, crashes, or fails healthcheck
- Prevents features from being permanently stuck after rate limit crashes
- Uses separate SQLAlchemy engine to avoid session conflicts with subprocess

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

server/services/expand_chat_session.py

@@ -154,16 +154,11 @@ class ExpandChatSession:
         system_prompt = skill_content.replace("$ARGUMENTS", project_path)
 
         # Build environment overrides for API configuration
-        # Filter to only include vars that are actually set (non-None)
-        sdk_env: dict[str, str] = {}
-        for var in API_ENV_VARS:
-            value = os.getenv(var)
-            if value:
-                sdk_env[var] = value
+        from registry import get_effective_sdk_env
+        sdk_env = get_effective_sdk_env()
 
-        # Determine model from environment or use default
-        # This allows using alternative APIs (e.g., GLM via z.ai) that may not support Claude model names
-        model = os.getenv("ANTHROPIC_DEFAULT_OPUS_MODEL", "claude-opus-4-5-20251101")
+        # Determine model from SDK env (provider-aware) or fallback to env/default
+        model = sdk_env.get("ANTHROPIC_DEFAULT_OPUS_MODEL") or os.getenv("ANTHROPIC_DEFAULT_OPUS_MODEL", "claude-opus-4-5-20251101")
 
         # Build MCP servers config for feature creation
         mcp_servers = {