feat: add Docker entrypoints, LLM providers, pipeline hardening, workbench pages

Phase 9 — four parallel workstreams:

- Stream A: 14 Docker entrypoints for containerized deployment
- Stream B: Pipeline hardening — robust JSON parsing, LLM retry logic,
  consumer negative-ack, FalkorDB test import fix
- Stream C: Azure OpenAI, OpenAI-compatible, and Mistral LLM providers
- Stream D: Workbench Prompts, Token Cost, Knowledge Cores pages +
  Settings feature switches

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

ts/scripts/run-llm-mistral.ts

@@ -0,0 +1,16 @@
+/**
+ * Start the Mistral text-completion service.
+ *
+ * Usage: MISTRAL_TOKEN=... pnpm tsx scripts/run-llm-mistral.ts
+ *
+ * Env:
+ *   NATS_URL      (default: nats://localhost:4222)
+ *   MISTRAL_TOKEN (required)
+ *   MISTRAL_MODEL (default: ministral-8b-latest)
+ */
+import { run } from "../packages/flow/src/model/text-completion/mistral.js";
+
+run().catch((err) => {
+  console.error("Mistral LLM service failed:", err);
+  process.exit(1);
+});