feat(llm): add LLM integration module with OpenRouter and Claude Code adapters

Implements markitect/llm/ package with concrete LLMAdapter implementations:
- OpenRouterAdapter: HTTP via urllib with retry/backoff on 429/5xx
- ClaudeCodeAdapter: subprocess-based Claude CLI with stdin piping
- Factory pattern: create_adapter("openrouter") or create_adapter("claude-code")
- API key resolution chain: constructor > env var > project-root key file
- 42 unit tests, 2 integration tests (gated on API key / CLI availability)

Also adds the infospace-with-history example with Wealth of Nations VSM
analysis pipeline, templates, schemas, source chapters, and processed
output for chapters 1-2. process_chapters.py now supports --provider
and --model flags for automatic LLM-driven processing.

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

tests/unit/llm/test_openrouter.py

@@ -0,0 +1,143 @@
+"""Tests for markitect.llm.openrouter."""
+
+from unittest import mock
+
+import pytest
+
+from markitect.llm.openrouter import OpenRouterAdapter
+from markitect.llm.config import LLMConfig
+from markitect.llm.exceptions import LLMRateLimitError, LLMAPIError
+from markitect.prompts.execution.models import RunConfig, LLMResponse
+
+
+def _make_api_response(content="hello world", model="test-model"):
+    return {
+        "id": "gen-abc123",
+        "model": model,
+        "choices": [
+            {
+                "message": {"role": "assistant", "content": content},
+                "finish_reason": "stop",
+            }
+        ],
+        "usage": {
+            "prompt_tokens": 10,
+            "completion_tokens": 5,
+            "total_tokens": 15,
+        },
+    }
+
+
+class TestOpenRouterAdapter:
+    def _adapter(self, **kwargs):
+        defaults = {"api_key": "sk-test", "model": "test/model"}
+        defaults.update(kwargs)
+        return OpenRouterAdapter(**defaults)
+
+    @mock.patch("markitect.llm.openrouter.post_json")
+    def test_execute_prompt_success(self, mock_post):
+        mock_post.return_value = _make_api_response("Generated text")
+        adapter = self._adapter()
+        config = RunConfig(model_name="test/model", temperature=0.5, max_tokens=100)
+
+        resp = adapter.execute_prompt("Write something", config)
+
+        assert isinstance(resp, LLMResponse)
+        assert resp.content == "Generated text"
+        assert resp.usage["prompt_tokens"] == 10
+        assert resp.usage["completion_tokens"] == 5
+        assert resp.metadata["provider"] == "openrouter"
+        assert "latency_seconds" in resp.metadata
+
+    @mock.patch("markitect.llm.openrouter.post_json")
+    def test_payload_structure(self, mock_post):
+        mock_post.return_value = _make_api_response()
+        adapter = self._adapter(system_prompt="You are helpful")
+        config = RunConfig(temperature=0.3, max_tokens=500)
+
+        adapter.execute_prompt("Test prompt", config)
+
+        call_args = mock_post.call_args
+        payload = call_args[0][1]  # second positional arg
+        assert payload["model"] == "test/model"
+        assert payload["temperature"] == 0.3
+        assert payload["max_tokens"] == 500
+        assert len(payload["messages"]) == 2
+        assert payload["messages"][0]["role"] == "system"
+        assert payload["messages"][1]["role"] == "user"
+
+    @mock.patch("markitect.llm.openrouter.post_json")
+    def test_no_system_prompt(self, mock_post):
+        mock_post.return_value = _make_api_response()
+        adapter = self._adapter()
+        config = RunConfig()
+
+        adapter.execute_prompt("Test", config)
+
+        payload = mock_post.call_args[0][1]
+        assert len(payload["messages"]) == 1
+        assert payload["messages"][0]["role"] == "user"
+
+    @mock.patch("markitect.llm.openrouter.post_json")
+    @mock.patch("markitect.llm.openrouter.time.sleep")
+    def test_retry_on_429(self, mock_sleep, mock_post):
+        mock_post.side_effect = [
+            LLMRateLimitError("rate limited", status_code=429),
+            _make_api_response("after retry"),
+        ]
+        adapter = self._adapter(max_retries=2)
+        config = RunConfig()
+
+        resp = adapter.execute_prompt("Test", config)
+        assert resp.content == "after retry"
+        assert mock_sleep.call_count == 1
+
+    @mock.patch("markitect.llm.openrouter.post_json")
+    @mock.patch("markitect.llm.openrouter.time.sleep")
+    def test_retry_on_5xx(self, mock_sleep, mock_post):
+        mock_post.side_effect = [
+            LLMAPIError("server error", status_code=502),
+            _make_api_response("recovered"),
+        ]
+        adapter = self._adapter(max_retries=2)
+        config = RunConfig()
+
+        resp = adapter.execute_prompt("Test", config)
+        assert resp.content == "recovered"
+
+    @mock.patch("markitect.llm.openrouter.post_json")
+    def test_no_retry_on_4xx(self, mock_post):
+        mock_post.side_effect = LLMAPIError("bad request", status_code=400)
+        adapter = self._adapter(max_retries=2)
+        config = RunConfig()
+
+        with pytest.raises(LLMAPIError) as exc_info:
+            adapter.execute_prompt("Test", config)
+        assert exc_info.value.status_code == 400
+
+    @mock.patch("markitect.llm.openrouter.post_json")
+    @mock.patch("markitect.llm.openrouter.time.sleep")
+    def test_exhausted_retries_raises(self, mock_sleep, mock_post):
+        mock_post.side_effect = LLMRateLimitError("rate limited", status_code=429)
+        adapter = self._adapter(max_retries=1)
+        config = RunConfig()
+
+        with pytest.raises(LLMRateLimitError):
+            adapter.execute_prompt("Test", config)
+
+    def test_validate_config_no_key(self):
+        adapter = OpenRouterAdapter(api_key=None, model="test")
+        # Force key to None
+        adapter._api_key = None
+        config = RunConfig()
+        assert adapter.validate_config(config) is False
+
+    def test_validate_config_ok(self):
+        adapter = self._adapter()
+        config = RunConfig(temperature=0.5)
+        assert adapter.validate_config(config) is True
+
+    def test_validate_config_bad_temperature(self):
+        adapter = self._adapter()
+        config = RunConfig(temperature=3.0)
+        assert adapter.validate_config(config) is False