feat: major architectural refactor

- Refactor memory system (episodic/STM/LTM with components)
- Implement complete subtitle domain (scanner, matcher, placer)
- Add YAML workflow infrastructure
- Externalize knowledge base (patterns, release groups)
- Add comprehensive testing suite
- Create manual testing CLIs

tests/test_tools_language.py

@@ -0,0 +1,41 @@
+"""Tests for language tools."""
+
+import pytest
+
+from alfred.agent.tools.language import set_language
+
+
+class TestSetLanguage:
+
+    def test_success_returns_ok(self, memory):
+        result = set_language("fr")
+        assert result["status"] == "ok"
+        assert result["language"] == "fr"
+
+    def test_message_contains_language(self, memory):
+        result = set_language("en")
+        assert "en" in result["message"]
+
+    def test_persists_to_memory(self, memory):
+        set_language("es")
+        # Verify it's stored in STM
+        from alfred.infrastructure.persistence import get_memory
+        mem = get_memory()
+        assert mem.stm.language == "es"
+
+    def test_various_language_codes(self, memory):
+        for lang in ["en", "fr", "es", "de", "it", "pt"]:
+            result = set_language(lang)
+            assert result["status"] == "ok"
+            assert result["language"] == lang
+
+    def test_error_on_memory_failure(self, monkeypatch):
+        from alfred.agent.tools import language as lang_module
+
+        def broken_get_memory():
+            raise RuntimeError("memory unavailable")
+
+        monkeypatch.setattr(lang_module, "get_memory", broken_get_memory)
+        result = set_language("fr")
+        assert result["status"] == "error"
+        assert "error" in result