Keep a record of each run's tool calls so bots can't claim work they never did

The wrap-up turn now gets the run's actual tool calls, each bot message
keeps them as bot_messages.tool_log, and history follows recorded bot
messages with an [App record] line. The coordinator also gets update_bot
in workspaces, so it can fill a bot's core prompt from there.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

src/harness/text/describe-tool-uses.test.ts

@@ -0,0 +1,22 @@
+import { describe, expect, test } from "bun:test";
+import describeToolUses from "./describe-tool-uses";
+
+describe("describeToolUses", () => {
+    test("says none when no tool ran", () => {
+        expect(describeToolUses({ uses: [] })).toBe("none");
+    });
+
+    test("groups calls by tool, in the order first called, and marks failures", () => {
+        expect(
+            describeToolUses({
+                uses: [
+                    { name: "fetch_url", ok: true },
+                    { name: "update_memory", ok: true },
+                    { name: "fetch_url", ok: false },
+                    { name: "fetch_url", ok: true },
+                    { name: "set_bot_workspace_instructions", ok: false },
+                ],
+            }),
+        ).toBe("fetch_url x3 (1 failed), update_memory, set_bot_workspace_instructions (failed)");
+    });
+});