feat: sync runtime search and schema quality updates from app repo

- port retrieval, validation, and eval improvements relevant to os
- align prompts and dimensions with the flat single-agent model
- replace the old eval suite with the focused core scenarios

Generated with Codex

tests/evals/scenarios/archive/hard-mode-query.ts

@@ -0,0 +1,16 @@
+import { Scenario } from '../types';
+
+export const scenario: Scenario = {
+  id: 'hard-mode-query',
+  name: 'Hard mode retrieval query',
+  description: 'Run a baseline retrieval query in hard mode.',
+  tools: ['queryNodes', 'searchContentEmbeddings'],
+  input: {
+    message: 'What have I captured about plaintext productivity and tools?',
+    mode: 'hard',
+  },
+  expect: {
+    toolsCalledSoft: ['queryNodes'],
+    responseContainsSoft: ['plaintext', 'productivity'],
+  },
+};