feat(ai): wire llm_features into compliance, seo, reconciliation

The AI features in llm_features.py (llm_compliance_analyze,
llm_seo_analyze, llm_entity_reconcile, llm_pii_detect,
llm_anomaly_detect) were implemented but never called from the live
code path. The endpoint functions were regex-only, with the LLM
functions sitting in limbo.

This change wires the LLM as a FALLBACK when the regex/heuristic
pass is low-confidence. The user pays nothing extra, gets better
results, and the LLM cost is tracked per-call.

Changes:
- compliance.py run_compliance_check:
    When tos_result.confidence == "low" (or no ToS was found),
    call llm_compliance_analyze and merge the richer classification
    into tos_result. llm_enhanced: True is set.
    Pass-through: the LLM fields (provider, cost, risk_summary, etc.)
    are now copied into the terms_of_service sub-dict of the response.
- seo_monitor.py analyze_seo:
    When title, meta_description, or h1 are empty after the regex
    pass, call llm_seo_analyze to suggest content. Best-effort: empty
    regex fields are filled in from LLM suggestions, llm_enhanced
    flag is set.
- reconciliation.py:
    New async function llm_enhance_reconciliation(entities) that
    sends low-confidence groups to llm_entity_reconcile for
    verification/refutation. Returns a summary dict with counts.
- New test file tests/test_llm_fallback.py with 6 tests:
    compliance: 2 tests (merges correctly, degrades on LLM error)
    seo: 1 test (fills empty fields, sets llm_enhanced)
    reconciliation: 3 tests (function exists, handles no-low-conf,
      handles LLM error)
    All 6 pass. All existing compliance/seo/reconciliation tests
    (28) still pass.

Defaults: the LLM uses the fleet's free Ollama on Talos
(100.100.18.18:11434) when no other provider is configured, so
fallback cost is effectively zero in production.

seo_monitor.py

@@ -68,6 +68,29 @@ async def analyze_seo(url: str) -> dict[str, Any]:
             "content_type": resp.headers.get("content-type", ""),
             "last_modified": resp.headers.get("last-modified", ""),
         }
+        # LLM enhancement: if critical SEO fields are empty, use the LLM to
+        # suggest better content based on the page. Best-effort: if the LLM
+        # call fails or no provider is configured, we return the regex result.
+        missing_critical = [f for f in ("title", "meta_description", "h1") if not result.get(f)]
+        if missing_critical:
+            try:
+                from llm_features import llm_seo_analyze
+                llm_enhancement = await llm_seo_analyze(
+                    url=url,
+                    html=resp.text[:6000],
+                    missing_fields=missing_critical,
+                )
+                if llm_enhancement:
+                    for f in missing_critical:
+                        suggestion = llm_enhancement.get(f)
+                        if suggestion and not result.get(f):
+                            result[f] = suggestion
+                    result["llm_enhanced"] = True
+                    result["llm_provider"] = llm_enhancement.get("llm_provider", "")
+                    result["llm_cost_usd"] = llm_enhancement.get("llm_cost_usd", 0.0)
+            except Exception as e:
+                logger.debug("llm_seo_enhance_failed", extra={"url": url, "error": str(e)[:80]})
+
         return result
     except Exception as e:
         return {"url": url, "error": str(e)[:200]}