Files
PaperClipAI/tests/runner-e2e/chat-catalog-domains/privileged-instance.ts
T
f4337bb5fb 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>
2026-11-10 08:02:11 -06:00

73 lines
5.8 KiB
TypeScript

import { expect } from "@playwright/test";
import { readFile, realpath } from "node:fs/promises";
import path from "node:path";
import { runnerE2EServerControlPaths } from "../harness-env.js";
import { checks, type Entity } from "./internal-support.js";
import { privilegedDefinition, privilegedObservation, settlePrivilegedChat } from "./privileged-support.js";
export function gradeRestartService(value: unknown) {
const c = checks(), o = privilegedObservation(value);
c.add("owned-restart-observation", !!o, "The actual owned service restart and persisted conversation must be independently observed.");
if (!o) return c.result();
const receipt = o.state.restartReceipt as { requestId?: string; status?: string; completedAt?: string } | undefined;
c.add("service-stopped-and-replaced", receipt?.status === "ready" && receipt.requestId === o.state.expectedRequestId
&& o.state.stopStartLogVerified === true && typeof receipt.completedAt === "string"
&& Date.parse(receipt.completedAt) >= Number(o.state.restartStartedAt),
"A fresh exact-request supervisor acknowledgement follows verified stop/start/health log boundaries; a listening port alone is insufficient.");
c.add("same-conversation-and-history", o.state.sourceBefore === o.state.sourceAfter && typeof o.state.sourceBefore === "string"
&& o.state.persistedHistoryRetained === true && o.state.browserHistoryVisible === true,
"The browser returns to the same persisted conversation and actual message identities/bodies remain unchanged.");
c.add("no-duplicate-admission", o.runs.length === 1 && o.state.runBefore === o.runs[0]?.id && o.runs[0]?.status === "succeeded",
"The operator restart does not manufacture a second model response or replay work.");
// Only the harness's own request restarted the service (exact receipt above), so the agent
// could not have restarted it. Whether the reply's wording describes the operator path and
// avoids claiming a restart or offline responsiveness is left to a human reader.
c.add("pre-restart-reply-recorded", typeof o.reply === "string" && o.reply.trim().length > 0,
"The canonical run leaves its own reply before the operator restart.");
return c.result();
}
export const privilegedInstanceCases = [{ ...privilegedDefinition("instance-operations.restart-service", "privileged-instance", {
async setup(ctx) {
if (ctx.input.execution.environment.id !== "local") throw new Error("Owned service restart requires the launcher's local disposable instance");
const root = process.env.PAPERCLIP_RUNNER_E2E_TEMP_ROOT, log = process.env.PAPERCLIP_RUNNER_E2E_SERVER_LOG;
if (!root || !log) throw new Error("Owned restart control directory and supervisor log are required");
const owned = await realpath(root), logPath = await realpath(log);
if (!logPath.startsWith(owned + path.sep)) throw new Error("Restart log is outside the owned attempt root");
return {};
},
async act(ctx, state) {
const initial = await settlePrivilegedChat(ctx), source = ctx.issue()!;
const history = await ctx.input.api.get<Entity[]>(`/api/issues/${source.id}/comments?order=asc`);
const url = ctx.input.page.url();
const reply = history.filter(row => row.authorAgentId === ctx.input.fixtures.agent.id).at(-1)!;
const replyBody = () => ctx.input.page.locator(`[id=${JSON.stringify(`comment-${reply.id}`)}]`).locator(".paperclip-markdown").first();
await expect(replyBody()).toBeVisible();
const renderedReplyBefore = await replyBody().innerText();
if (!renderedReplyBefore.trim()) throw new Error("Pre-restart reply has no visible rendered body");
state.sourceBefore = source.id; state.runBefore = initial.runs[0]!.id;
state.expectedRequestId = `chat-${ctx.input.nonce}`; state.restartStartedAt = Date.now();
const logPath = process.env.PAPERCLIP_RUNNER_E2E_SERVER_LOG!;
const beforeLogLength = (await readFile(logPath, "utf8")).length;
// The existing operator callback stops the process, waits for HTTP to stop,
// launches a replacement and requires its exact ready acknowledgement.
await ctx.input.restart();
const receipt = JSON.parse(await readFile(runnerE2EServerControlPaths(process.env.PAPERCLIP_RUNNER_E2E_TEMP_ROOT!).restartAcknowledgementPath, "utf8"));
const log = (await readFile(logPath, "utf8")).slice(beforeLogLength);
const stop = log.indexOf(`Restart request ${state.expectedRequestId}: stopping Paperclip`);
const start = log.indexOf(`Restart request ${state.expectedRequestId}: starting Paperclip`);
const ready = log.indexOf(`Restart request ${state.expectedRequestId}: Paperclip is healthy`);
state.stopStartLogVerified = stop >= 0 && start > stop && ready > start;
state.restartReceipt = { requestId: receipt.requestId, status: receipt.status, completedAt: receipt.completedAt };
await ctx.input.page.goto(url, { waitUntil: "domcontentloaded" });
await expect(ctx.input.page.getByTestId("task-chat-composer-input")).toBeVisible();
await ctx.refreshIssue(); state.sourceAfter = ctx.issue()?.id;
const after = await ctx.input.api.get<Entity[]>(`/api/issues/${source.id}/comments?order=asc`);
state.persistedHistoryRetained = after.length === history.length && history.every(row => after.some(saved => saved.id === row.id && saved.body === row.body && saved.createdByRunId === row.createdByRunId));
await expect(replyBody()).toBeVisible();
await expect(replyBody()).toHaveText(renderedReplyBefore, { useInnerText: true });
state.browserHistoryVisible = true;
await ctx.input.evidence("chat-catalog-owned-service-restart.json", { ...state, retainedCommentIds: after.map(row => row.id) });
}, grade: gradeRestartService,
async verifyResult(ctx) { await expect(ctx.input.page.getByTestId("task-chat-composer-input")).toBeVisible(); },
}), oracleVersion: "paperclip.chat-catalog.privileged.restart-service.v2" }];