Files
PaperClipAI/tests/runner-e2e/chat-catalog-domains/external-app-test.test.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

76 lines
7.9 KiB
TypeScript

import { expect, it, vi } from "vitest";
import type { ChatCatalogContext } from "../chat-catalog-contract.js";
const port = vi.hoisted(() => ({ calls: [] as Array<Record<string, unknown>>, close: vi.fn(async () => {}), probe: vi.fn(async () => 200), missing: false, reply: "Morgan's sample-page test succeeded." }));
vi.mock("./external-provider.js", () => ({ startNotesProvider: async () => ({ url: "http://127.0.0.1:492/", token: "synthetic-private-app492-credential", calls: port.calls, close: port.close }), connectNotes: async () => ({ connectionId: "connection", read: ["read"], write: ["write"] }) }));
vi.mock("./identity-support.js", () => ({ identityWorker: async () => ({ id: "morgan" }), waitForReply: vi.fn(async () => {}), cleanupIdentityProbes: vi.fn(async () => {}), probeAgentPermission: port.probe }));
vi.mock("./external-support.js", async original => ({ ...await original<typeof import("./external-support.js")>(), observeExternal: async (ctx: ChatCatalogContext, state: Record<string, unknown>) => ({
scenarioId: "app-permissions.test-as-agent", origin: "https://fixture.test", companyId: "company", leadId: "ceo", state,
tasks: [], projects: [], goals: [], comments: [], documents: [], revisions: [], activity: [], humanQueries: [], links: [], reply: port.reply,
runs: state.sourceIssueId ? [{ id: "native", companyId: "company", agentId: "ceo", runtimeMode: "native", status: "succeeded", contextSnapshot: { issueId: "source", commentId: "human" } }] : [],
phaseReplies: [], connections: [], profiles: [], secrets: [], endpoints: [], interactions: [],
}) }));
import { appTestExternalCases } from "./external-app-test.js";
const definition = appTestExternalCases[0]!;
function fixture() {
port.calls.splice(0); port.close.mockClear(); port.probe.mockClear(); port.missing = false; port.reply = "Morgan's sample-page test succeeded.";
const metadata = { source: "test", headerSummary: { value: "synthetic-private-metadata-canary" }, execution: { request: { endpoint: "http://127.0.0.1:492/", mcpMethod: "tools/call", upstreamToolName: "launch_read", requestId: "paperclip-tool-owned" }, unknownToken: "synthetic-private-metadata-canary" } };
const get = vi.fn(async (path: string) => {
if (path === "/api/auth/get-session") return { user: { id: "person" } };
if (path.includes("/profiles/effective/agents/")) return { agentId: path.split("/").at(-1), allowedTools: port.missing ? [] : [{ id: "read" }], profiles: [], entries: [], bindings: [], installedConnections: [{ id: "connection" }] };
if (path.endsWith("/runs/native/decisions")) return { runId: "native", decisions: [{ invocation: { id: "invocation", companyId: "company", actorType: "agent", actorId: "ceo", agentId: "morgan", issueId: "source", runId: "native", connectionId: "connection", catalogEntryId: "read", toolName: "qualified-read", status: "succeeded", policyDecision: "allow", errorCode: null, resultHash: "a".repeat(64), resultSizeBytes: 100, startedAt: "2026-11-03T00:00:01Z", completedAt: "2026-11-03T00:00:02Z" }, auditEvents: [{ id: "event", invocationId: "invocation", eventType: "call_completed", toolName: "qualified-read", metadata }] }] };
if (path.includes("/comments?")) return [{ id: "human", authorAgentId: null, authorUserId: "person", createdByRunId: null, body: "Check whether Morgan can read the sample page with this connection.", createdAt: "2026-11-03T00:00:00Z" }];
if (path.includes("/heartbeat-runs?")) return [];
if (path === "/api/tool-connections/connection/test-agents/morgan/access") return { access: { connectionId: "connection", toolCount: 2, allowedCount: 1, askFirstCount: 0, offCount: 1, tools: [{ toolName: "launch_read", gatewayToolName: "qualified-read", risk: "read", decision: "allowed", reasonCode: "allow_profile", matchedPolicyIds: [] }, { toolName: "launch_update", gatewayToolName: "qualified-write", risk: "destructive", decision: "off", reasonCode: "deny_default", matchedPolicyIds: [] }] } };
throw Error("Unexpected owned read");
});
const api = { get, putSensitive: vi.fn(async () => {}), post: vi.fn(async () => {}), postSensitive: vi.fn(async () => { throw Error("No pre-query test is permitted"); }), patch: vi.fn(async () => {}), delete: vi.fn(async () => {}) };
const ctx = { input: { api, fixtures: { company: { id: "company", issuePrefix: "UNIT" }, agent: { id: "ceo" } } }, scenario: { query: "Check whether Morgan can read the sample page with this connection." }, issue: () => ({ id: "source" }), refreshIssue: vi.fn(async () => {}) } as unknown as ChatCatalogContext;
return { ctx, api, metadata };
}
it("actual setup preserves real read grants and never probes or pre-completes the requested test", async () => {
const f = fixture(); let state: Record<string, unknown> = {};
try {
state = await definition.setup(f.ctx);
expect(port.probe).not.toHaveBeenCalled(); expect(f.api.postSensitive).not.toHaveBeenCalled();
expect(state).toMatchObject({ providerCallsBeforeQuery: 0, initialCeoScope: { agentId: "ceo", allowed: ["read"] }, initialMorganScope: { agentId: "morgan", allowed: ["read"] } });
expect(f.api.post).toHaveBeenCalledTimes(1); expect(f.api.putSensitive).toHaveBeenCalledTimes(1);
} finally { await definition.cleanup!(f.ctx, state); }
expect(port.close).toHaveBeenCalledTimes(1);
});
it("refuses a denied initial grant instead of supplying the outcome or granting more access", async () => {
const f = fixture(); port.missing = true;
await expect(definition.setup(f.ctx)).rejects.toThrow("scoped access must already allow");
expect(f.api.postSensitive).not.toHaveBeenCalled(); expect(f.api.post).not.toHaveBeenCalled(); expect(port.close).toHaveBeenCalledTimes(1);
});
it("actual observer reads the canonical run's saved decision and retains only closed test provenance", async () => {
const f = fixture(); let state: Record<string, unknown> = {};
try {
state = await definition.setup(f.ctx);
port.calls.push({ method: "tools/call", name: "launch_read", authorized: true, arguments: { hidden: "synthetic-private-metadata-canary" }, responseHttpStatus: 200, receivedAt: 1, respondedAt: 2 });
await definition.observe(f.ctx, state);
expect(f.api.get).toHaveBeenCalledWith("/api/companies/company/tools/runs/native/decisions");
expect(state).toMatchObject({ decisions: [{ runId: "native", invocation: { agentId: "morgan", actorId: "ceo" }, events: [{ source: "test", endpointMatches: true, method: "tools/call", upstreamToolName: "launch_read", requestIdSha256: expect.stringMatching(/^[a-f0-9]{64}$/) }] }], credentialDisclosed: false, morganRuntimeRunCount: 0 });
expect(JSON.stringify(state)).not.toContain("synthetic-private-metadata-canary");
expect(JSON.stringify(state)).not.toContain("synthetic-private-app492-credential");
expect(port.probe).not.toHaveBeenCalled(); expect(f.api.postSensitive).not.toHaveBeenCalled();
} finally { await definition.cleanup!(f.ctx, state); }
});
it("actual observer leaves unknown audit source and request identity ineligible", async () => {
const f = fixture(); let state: Record<string, unknown> = {};
try {
state = await definition.setup(f.ctx); f.metadata.source = "unrelated"; f.metadata.execution.request.requestId = "unknown request with spaces";
await definition.observe(f.ctx, state);
expect(state).toMatchObject({ decisions: [{ events: [{ source: "other", requestIdSha256: null }] }] });
} finally { await definition.cleanup!(f.ctx, state); }
});
it("actual observer flags an encoded credential disclosure without retaining its value", async () => {
const f = fixture(); let state: Record<string, unknown> = {};
try {
state = await definition.setup(f.ctx);
port.reply += Buffer.from("synthetic-private-app492-credential").toString("base64");
await definition.observe(f.ctx, state);
expect(state.credentialDisclosed).toBe(true);
expect(JSON.stringify(state)).not.toContain("synthetic-private-app492-credential");
} finally { await definition.cleanup!(f.ctx, state); }
});