Files
PaperClipAI/tests/runner-e2e/chat-catalog-domains/runtime-support.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

111 lines
7.9 KiB
TypeScript

import { expect, type Page } from "@playwright/test";
import { pollUntil } from "../api.js";
import type { ChatCatalogCase, ChatCatalogContext } from "../chat-catalog-contract.js";
import { codexSubscriptionAgentConfig } from "../codex-subscription-fixture.js";
import { observeInternal, type Entity, type InternalObservation } from "./internal-support.js";
export interface RuntimeAgent extends Entity {
reportsTo?: string | null; adapterType?: string; adapterConfig?: Record<string, unknown>;
runtimeConfig?: Record<string, unknown>; permissions?: Record<string, unknown>;
defaultEnvironmentId?: string | null; lifecycleState?: string; lifecycleHolds?: string[]; pauseReason?: string | null;
}
export interface InstructionFile {
path: string; content: string; revision: { id: string; parentRevisionId?: string | null;
companyId: string; agentId: string; actorAgentId?: string | null; sourceRunId?: string | null; source: string; restoredFromRevisionId?: string | null };
}
export interface RuntimeObservation extends InternalObservation {
agents: RuntimeAgent[];
instruction?: InstructionFile;
instructionHistory?: { revisions: InstructionFile["revision"][] };
instructionSideFile?: { content: string };
}
export async function runtimeWorker(ctx: ChatCatalogContext, name = "Morgan", extra: Record<string, unknown> = {}) {
const { api, execution, fixtures, workspacePath, nonce } = ctx.input;
const base = execution.profile.buildAgent({ environmentId: fixtures.environment.id, environmentFixtureId: execution.environment.id,
workspacePath, secretRefs: fixtures.secretRefs, executionId: nonce });
const payload = { ...base, name, role: "engineer", title: "Frontend engineer", reportsTo: fixtures.agent.id,
runtimeConfig: { ...(base.runtimeConfig as Record<string, unknown>), heartbeat: { enabled: false, maxConcurrentRuns: 1 } }, ...extra };
const agent = await api.post<RuntimeAgent>(`/api/companies/${fixtures.company.id}/agents`, execution.profile.credential === "CODEX_AUTH_JSON"
? codexSubscriptionAgentConfig(payload, process.env, fixtures.company.id) : payload);
return pollUntil({ label: `${name} real runtime readiness`, deadlineAt: Date.now() + 45_000,
load: () => api.get<RuntimeAgent>(`/api/agents/${agent.id}`), accept: row => row.lifecycleState === "ready" });
}
export async function observeRuntime(ctx: ChatCatalogContext, state: Record<string, unknown>): Promise<RuntimeObservation> {
const o = await observeInternal(ctx, state);
const agents = await ctx.input.api.get<RuntimeAgent[]>(`/api/companies/${o.companyId}/agents`);
const result: RuntimeObservation = { ...o, agents };
if (state.instructionAgentId) {
const root = `/api/agents/${state.instructionAgentId}/instructions-bundle`;
result.instruction = await ctx.input.api.get<InstructionFile>(`${root}/file?path=AGENTS.md`);
result.instructionHistory = await ctx.input.api.get(`${root}/history?path=AGENTS.md`);
result.instructionSideFile = await ctx.input.api.get(`${root}/file?path=guidance.md`);
}
return result;
}
export function runtimeObservation(value: unknown): RuntimeObservation | undefined {
const o = value as RuntimeObservation | undefined;
return o && o.state && typeof o.reply === "string" && typeof o.companyId === "string" && typeof o.leadId === "string"
&& [o.agents, o.tasks, o.links, o.activity, o.runs, o.humanQueries].every(Array.isArray) ? o : undefined;
}
export function agentLink(o: RuntimeObservation, agent: RuntimeAgent): string | undefined {
return o.links.find(href => {
try { const url = new URL(href, o.origin); const parts = url.pathname.split("/").map(decodeURIComponent); const index = parts.indexOf("agents");
return url.origin === new URL(o.origin).origin && !url.pathname.startsWith("/api/") && index >= 0 && [agent.id, agent.urlKey].includes(parts[index + 1]); }
catch { return false; }
});
}
/** Canonical outcomes may name a supported control without a Markdown href. */
export async function openRuntimeAgentProfile(ctx: ChatCatalogContext, agent: RuntimeAgent, o?: RuntimeObservation) {
const href = o && agentLink(o, agent);
if (href) await ctx.input.page.locator(`[id^="comment-"] a[href=${JSON.stringify(href)}]`).first().click();
else await ctx.input.page.goto(`/${ctx.input.fixtures.company.issuePrefix}/agents/${agent.id}`, { waitUntil: "domcontentloaded" });
await expect(agentProfileIdentity(ctx.input.page, agent)).toBeVisible();
}
/**
* The profile header's h1 is the agent's editable title (falling back to its
* role), and the name is secondary text. The header portrait carries the
* name as "<name> avatar", so it identifies whose profile is open.
*/
export function agentProfileIdentity(page: Page, agent: RuntimeAgent) {
return page.locator("header").getByRole("img", { name: `${agent.name} avatar`, exact: true }).first();
}
export async function verifyRuntimeResult(ctx: ChatCatalogContext, state: Record<string, unknown>, value: unknown) {
const o = runtimeObservation(value); if (!o) throw new Error("Missing independent runtime result");
const target = o.agents.find(row => row.id === (state.worker as RuntimeAgent | undefined)?.id);
if (!target) { await expect(ctx.input.page.getByTestId("task-chat-composer-input")).toBeVisible(); return; }
const source = ctx.input.page.url();
await openRuntimeAgentProfile(ctx, target, o);
await ctx.input.page.reload({ waitUntil: "domcontentloaded" });
await expect(agentProfileIdentity(ctx.input.page, target)).toBeVisible();
await ctx.input.evidence("chat-catalog-runtime-profile-opened.json", { agentId: target.id, companyId: target.companyId, url: ctx.input.page.url() });
await ctx.input.page.goto(source, { waitUntil: "domcontentloaded" });
}
/**
* A pending interaction renders inline in the thread as `#interaction-<id>`,
* or docked in the composer takeover, which carries its title but no id.
*/
export function interactionSurface(page: Page, interaction: { id: string; title?: string | null }) {
const inline = page.locator(`[id=${JSON.stringify(`interaction-${interaction.id}`)}]`);
return interaction.title
? inline.or(page.getByTestId("task-chat-composer-takeover").filter({ hasText: interaction.title }))
: inline;
}
/** Open a pending request from its timeline chip when it is not already docked, then use its own action. */
export async function clickInteractionAction(page: Page, interaction: { id: string; title?: string | null }, name: string | RegExp) {
const docked = page.getByTestId("task-chat-composer-takeover").filter({ hasText: interaction.title ?? "" });
const action = (interaction.title ? docked : interactionSurface(page, interaction)).getByRole("button", typeof name === "string" ? { name, exact: true } : { name });
await expect(interactionSurface(page, interaction).first()).toBeVisible();
if (!(await action.first().isVisible())) await page.locator(`[id=${JSON.stringify(`interaction-${interaction.id}`)}]`).click();
await action.first().click();
}
export function runtimeDefinition(scenarioId: string, setup: ChatCatalogCase["setup"], grade: ChatCatalogCase["grade"], extra: Partial<ChatCatalogCase> = {}): ChatCatalogCase {
return { scenarioId, setup, grade, observe: observeRuntime, verifyResult: verifyRuntimeResult,
oracleVersion: "paperclip.chat-catalog.runtime.v2", coverageGaps: [], expectedRunCount: 1,
turnTimeoutMs: 120_000, attemptTimeoutMs: 300_000,
sourceFiles: ["chat-catalog-domains/runtime-support.ts", `chat-catalog-domains/runtime-${scenarioId.split(".")[0]}.ts`, `chat-scenarios/${scenarioId.split(".")[0]}.yaml`], ...extra };
}
export function unchangedAgent(before: RuntimeAgent, after: RuntimeAgent, except: readonly string[] = []) {
return ["id", "companyId", "name", "role", "adapterType", "adapterConfig", "runtimeConfig", "permissions", "defaultEnvironmentId", "reportsTo"]
.filter(key => !except.includes(key)).every(key => JSON.stringify(before[key as keyof RuntimeAgent]) === JSON.stringify(after[key as keyof RuntimeAgent]));
}