feat(brain): add brain_answer and brain_classify MCP tools

Adds two new LLM-backed MCP tools to the ingestion service:

- brain_answer(query): BM25 retrieval + LLM synthesis → answer + sources
- brain_classify(text): classifies doc into type/title/tags via LLM

Adds llm.Router for primary→fallback routing (berget.ai → iguana).
Wired via BRAIN_LLM_PRIMARY_URL/BRAIN_LLM_FALLBACK_URL env vars;
no-op when unset so existing deployments are unaffected.

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

ingestion/internal/llm/router.go

@@ -0,0 +1,29 @@
+package llm
+
+import (
+	"context"
+	"fmt"
+)
+
+// Router calls Primary first; on any error falls back to Fallback.
+// Fallback may be nil, in which case primary errors are returned directly.
+type Router struct {
+	Primary  *Client
+	Fallback *Client
+}
+
+// Complete implements pipeline.CompleteFunc, routing through Primary then Fallback.
+func (r *Router) Complete(ctx context.Context, system, user string) (string, error) {
+	out, err := r.Primary.Complete(ctx, system, user)
+	if err == nil {
+		return out, nil
+	}
+	if r.Fallback == nil {
+		return "", fmt.Errorf("primary llm: %w", err)
+	}
+	out, err2 := r.Fallback.Complete(ctx, system, user)
+	if err2 != nil {
+		return "", fmt.Errorf("primary llm: %w; fallback llm: %v", err, err2)
+	}
+	return out, nil
+}