feat(llm): cache hint TTL, breakpoint cap, and tool placement (#26779)

packages/llm/test/recorded-scenarios.ts

@@ -6,6 +6,19 @@ import { tool } from "../src/tool"
 
 export const weatherToolName = "get_weather"
 
+// A deterministic system prompt long enough to clear every supported provider's
+// minimum cacheable-prefix threshold (Anthropic Haiku 3.5: 2048 tokens; Anthropic
+// Opus/Haiku 4.5: 4096 tokens; OpenAI/Gemini/Bedrock: lower). Built by repeating
+// a fixed sentence — the cassette replays bit-for-bit, so the exact text matters
+// only when re-recording with `RECORD=true`.
+export const LARGE_CACHEABLE_SYSTEM = (() => {
+  const sentence =
+    "You are a concise, factual assistant. Answer precisely and avoid filler. Cite numbers when known. "
+  // ~100 chars per sentence × 250 repeats ≈ 25,000 chars ≈ 5k+ tokens, safely
+  // above every provider's threshold.
+  return sentence.repeat(250)
+})()
+
 export const weatherTool = LLM.toolDefinition({
   name: weatherToolName,
   description: "Get current weather for a city.",