Add Phase 5: ReAct-style agent loop with tool execution

Implement the core autonomy layer — AgentLoop streams LLM responses,
parses tool calls, executes them with permission checks, feeds results
back, and repeats until the task completes or finish is called.

- Add FinishTool for explicit loop termination
- Add tools parameter to LLMClient.stream_chat() for function calling
- Add compact tool result display (status line, not full output)
- Refactor REPL to delegate to AgentLoop.run_turn()
- Fix Ollama null content rejection (always send content as string)
- Add finish to auto_approve permissions
- 9 unit tests for agent loop (34 total, zero regressions)

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

app/models/message.py

@@ -35,8 +35,9 @@ class Message(BaseModel):
         """
         data: dict[str, Any] = {"role": self.role}
 
-        if self.content is not None:
-            data["content"] = self.content
+        # Ollama requires content to be a string, never null/missing —
+        # even on assistant messages that only contain tool_calls.
+        data["content"] = self.content or ""
 
         if self.tool_calls is not None:
             data["tool_calls"] = [tc.model_dump() for tc in self.tool_calls]