test(evals): land the chat catalog eval harness on master (#19614)

## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work
> - The chat-feature evals measure whether agents complete real chat
requests through the product
> - The chat catalog eval harness and its repairs live on the
`chat-eval/master-candidate` branch, not on master
> - Today's work made 139 of 242 chat evals pass in verified single-case
live runs. Each product fix landed on master as its own pull request
> - The eval-side changes (setups, graders, fault seams, evidence
capture) still exist only on the candidate
> - This pull request lands the harness on master. Master stays the
source of truth for product behavior
> - The benefit is one tree: evals run against master commits, and
product fixes reach the evals as soon as they merge

## Linked Issues or Issue Description

Refs #19441, Refs #19599, Refs #20270, Refs #20334, Refs #20342, Refs
#20344, Refs #20354, Refs #20360, Refs #20373, Refs #20382, Refs #20409

**What existing behavior does this improve?**
The chat catalog eval harness and the product it evaluates live on a
long-running branch outside master. Eval fixes do not reach master, so
master-based eval runs use old setups and graders.

**Subsystem affected**
`tests/runner-e2e` chat catalog harness, plus a small set of product
changes that exist only on the candidate.

**Proposed change**
Land the harness on master. Every product fix from this eval work
already merged to master as its own pull request. Conflicts keep
master's product behavior.

## What Changed

- Most of the diff (808 of 851 files) is under `tests/runner-e2e/`:
chat-catalog domains, scenarios, descriptors, the TestOps manifest,
private evidence tooling, and their tests.
- Graders now check saved state, setup literals and rendered or clicked
links. Many keyword and regex checks on free-form replies are removed.
Each changed grader has a bumped `oracleVersion` and calibration tests
(a valid reply, a reworded valid reply, and broken variants that must
fail).
- The shared link helper `hasLink` no longer treats a list link (for
example `/COM/issues?view=mine`) as a link to every task without a URL
key.
- Setup fixes build real state through supported product routes: routine
history, agent grants through the board grant route, budget holds with a
real cap, and a storage-layer archive fault for
`inbox-search.archive-partial-failure`.
- Browser boundary journeys (sign-in, chat-disabled, onboarding) now
write the standard pass evidence files.
- Retargeted cases, approved by the maintainer:
`agent-chat.chat-disabled` (Conference Room gate),
`agents.restore-instructions` (no saved instruction history),
`access.manage-membership` (member permission edits stay board-only).
- The private evidence trusted files (`manual-private-handoff.ts`,
`manual-private-retention.ts`, `manual-private-report-assets.ts`) are
byte-identical to master.
- Product files outside `tests/` (43 files, +1,328 −49) carry
candidate-only changes that master does not yet have. Most are tests.
The main code changes are in `request-human-input.ts`,
`heartbeat-run-summary.ts`, `native-completion-feedback.ts`,
`native-run-finalizer.ts`, `openapi.ts`, `secrets.ts` and
`IssueChatLongThreadPerf.tsx`.

## Verification

- `pnpm vitest run chat-catalog-domains chat-catalog-contract.test.ts
chat-catalog-consistency.test.ts unit-launcher.test.mjs
--testTimeout=180000` in `tests/runner-e2e`: 269 files pass, 3,550 tests
pass.
- `node --test chat-full-catalog-testops-manifest.test.mjs`: 10 pass.
- `node tests/runner-e2e/source-sdk-public-inputs.mjs <HEAD> --check`:
no change.
- `tsc` for `server` and `tests/runner-e2e`: clean.
- Live single-case verification: 139 of 242 cases passed on paid runs,
and a person read the saved evidence for each. These are single-case
runs, not one full-cohort run.

## Risks

- Large diff. Most of it is test-only, but the 43 non-test files change
product code paths. Review those files first.
- Some graders are now less strict on reply wording. A person must read
replies for meaning on those cases.
- 42 cases need disposable outside accounts (Notion, Slack, GitHub,
Daytona and others) and stay blocked. `access.invite-teammate` stays
failing until local agent isolation lands, because the runner's durable
state keeps a live invite token.
- The paid workflow compares the trusted private-evidence files with
master. They are unchanged, so this risk is low.

## Model Used

Anthropic Claude Opus 5.5 (`claude-opus-5-5`), with tool use and code
execution, in Claude Code. Subagents used the same model.

## Checklist

- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [ ] All Paperclip CI gates are green
- [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Paperclip <noreply@paperclip.ing>

ui/src/components/task-chat/QuestionForm.catalog-navigation.test.tsx

@@ -0,0 +1,53 @@
+// @vitest-environment jsdom
+import { readFileSync } from "node:fs";
+import { dirname, resolve } from "node:path";
+import { fileURLToPath } from "node:url";
+import { act } from "react";
+import { createRoot, type Root } from "react-dom/client";
+import { afterEach, beforeEach, expect, it, vi } from "vitest";
+import type { PaperclipQuestionSet } from "@paperclipai/adapter-utils";
+import { QuestionForm } from "./QuestionForm";
+vi.mock("./TaskChatRichInput", () => ({ TaskChatRichInput: () => <textarea /> }));
+vi.mock("@/components/MarkdownBody", () => ({ MarkdownBody: ({ children }: { children: string }) => <div>{children}</div> }));
+function queryByRole(scope: HTMLElement, role: "button" | "radio", options: { name: string; exact: true }): HTMLElement | null {
+  return [...scope.querySelectorAll<HTMLElement>("button")].find(element =>
+    (element.getAttribute("role") ?? "button") === role
+      && (element.getAttribute("aria-label") ?? element.textContent?.trim()) === options.name) ?? null;
+}
+function getByRole(scope: HTMLElement, role: "button" | "radio", options: { name: string; exact: true }): HTMLElement {
+  const found = queryByRole(scope, role, options);
+  if (!found) throw new Error(`Missing rendered ${role} ${options.name}`);
+  return found;
+}
+const fixture = readFileSync(resolve(dirname(fileURLToPath(import.meta.url)), "../../../../tests/runner-e2e/chat-catalog-domains/identity-questions-and-approvals.ts"), "utf8");
+const previousSelector = fixture.match(/await form\.getByRole\("button", \{ name: "([^"]+)", exact: true \}\)\.click\(\);\s*state\.retainedAudience/);
+const set: PaperclipQuestionSet = { schema: "paperclip.question_set.v1", questions: [
+  { id: "audience", prompt: "Who is the audience?", required: true, answerMode: "single_select", options: [{ id: "researchers", label: "Researchers" }] },
+  { id: "tone", prompt: "What tone should we use?", required: true, answerMode: "single_select", options: [{ id: "clear", label: "Clear" }] },
+] };
+let host: HTMLDivElement, form: HTMLDivElement, root: Root;
+const submit = vi.fn();
+beforeEach(() => { localStorage.clear(); submit.mockReset(); host = document.createElement("div"); document.body.append(host); form = document.createElement("div"); form.dataset.testid = "task-chat-composer-takeover"; host.append(form); root = createRoot(form); });
+afterEach(async () => { await act(async () => root.unmount()); host.remove(); localStorage.clear(); });
+async function secondPage() {
+  await act(async () => root.render(<QuestionForm id="catalog-53" questionSet={set} onSubmit={submit} />));
+  await act(async () => getByRole(form, "button", { name: "Next", exact: true }).click());
+  expect(form.textContent).toContain("This question is required."); expect(form.textContent).not.toContain("What tone should we use?"); expect(submit).not.toHaveBeenCalled();
+  await act(async () => getByRole(form, "radio", { name: "Researchers", exact: true }).click());
+  await act(async () => getByRole(form, "button", { name: "Next", exact: true }).click());
+  expect(form.textContent).toContain("What tone should we use?");
+}
+it("uses the actual catalog consumer selector on rendered production pagination and retains the first answer", async () => {
+  await secondPage(); expect(previousSelector).not.toBeNull();
+  const previous = getByRole(form, "button", { name: previousSelector![1]!, exact: true });
+  expect((previous as HTMLButtonElement).disabled).toBe(false); await act(async () => previous.click());
+  expect(getByRole(form, "radio", { name: "Researchers", exact: true }).getAttribute("aria-checked")).toBe("true");
+  expect(form.textContent).not.toContain("What tone should we use?"); expect(submit).not.toHaveBeenCalled();
+});
+it("does not substitute wrong navigation names or another question's previous control", async () => {
+  await secondPage();
+  for (const name of ["Back", "Previous item"]) expect(queryByRole(form, "button", { name, exact: true })).toBeNull();
+  expect(getByRole(form, "button", { name: "Previous question", exact: true })).not.toBe(getByRole(form, "button", { name: "Next question", exact: true }));
+  const foreign = document.createElement("button"); foreign.setAttribute("aria-label", "Previous question"); host.append(foreign);
+  expect(getByRole(form, "button", { name: "Previous question", exact: true })).not.toBe(foreign); expect(submit).not.toHaveBeenCalled();
+});

ui/src/components/task-chat/QuestionForm.personal-secret-handoff.test.tsx

@@ -0,0 +1,24 @@
+// @vitest-environment jsdom
+import { act } from "react";
+import { createRoot } from "react-dom/client";
+import { expect, it, vi } from "vitest";
+import { QuestionForm } from "./QuestionForm";
+vi.mock("./TaskChatRichInput", () => ({ TaskChatRichInput: () => <textarea /> }));
+vi.mock("@/components/MarkdownBody", () => ({ MarkdownBody: ({ children }: { children: string }) => <div>{children}</div> }));
+it("the existing question form submits the exact saved-value choice only after a human selection and submit", async () => {
+  const host = document.createElement("div"); document.body.append(host); const root = createRoot(host), submit = vi.fn();
+  try {
+    await act(async () => root.render(<QuestionForm id="bound-question" questionSet={{ schema: "paperclip.question_set.v1", submitLabel: "Confirm value", questions: [
+      { id: "saved", prompt: "Have you saved your own value?", answerMode: "single_select", required: true, options: [
+        { id: "yes", label: "Yes", description: "My personal value is saved." }, { id: "later", label: "Not yet" },
+      ] },
+    ] }} onSubmit={submit} />));
+    expect(submit).not.toHaveBeenCalled();
+    expect(host.querySelector('[id="foreign-question-saved-yes"]')).toBeNull();
+    const option = host.querySelector<HTMLButtonElement>('[id="bound-question-saved-yes"]'); expect(option).not.toBeNull();
+    await act(async () => option!.click()); expect(submit).not.toHaveBeenCalled();
+    const button = [...host.querySelectorAll("button")].find(node => node.textContent === "Confirm value"); expect(button).toBeDefined();
+    await act(async () => button!.click());
+    expect(submit).toHaveBeenCalledExactlyOnceWith({ schema: "paperclip.question_response.v1", answers: { saved: { selectedOptionIds: ["yes"] } } });
+  } finally { await act(async () => root.unmount()); host.remove(); localStorage.clear(); }
+});

ui/src/components/task-chat/QuestionForm.test.tsx

@@ -46,6 +46,19 @@ describe("QuestionForm initial text", () => {
     await remount(refreshed); expect(input().value).toBe("Operator edit\n");
     expect(submit).not.toHaveBeenCalled();
   });
+  it("shows required validation without advancing or submitting after enabled Next", async () => {
+    await render({ schema: "paperclip.question_set.v1", questions: [
+      { id: "audience", prompt: "Who is the audience?", required: true, answerMode: "text" },
+      { id: "tone", prompt: "What tone should we use?", required: true, answerMode: "text" },
+    ] });
+    const next = Array.from(container.querySelectorAll("button")).find(button => button.textContent === "Next")!;
+    expect(next.disabled).toBe(false);
+    await act(async () => next.click());
+    expect(container.textContent).toContain("This question is required.");
+    expect(container.textContent).toContain("Who is the audience?");
+    expect(container.textContent).not.toContain("What tone should we use?");
+    expect(submit).not.toHaveBeenCalled();
+  });
   it("retains an explicitly cleared draft and required validation across reload", async () => {
     await render(); await edit(""); await remount();
     expect(input().value).toBe("");