Files
PaperClipAI/tests/runner-e2e/chat-catalog-domains/integrator-routine-invocation.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

97 lines
8.4 KiB
TypeScript

import { describe, expect, it } from "vitest";
import { gradeIntegratorRoutineInvocation } from "./integrator-routine-invocation.js";
function valid() {
const task = { id: "task", identifier: "ACN-2", projectId: "project", assigneeAgentId: "ceo", status: "done", title: "Launch summary" };
const invocation = { id: "invocation", routineId: "routine", source: "manual", status: "completed", linkedIssueId: "task", idempotencyKey: "delivery",
completedAt: "2026-01-01T00:02:00.000Z", triggerPayload: { variables: { projectName: "Launch" } } };
const chatRun = { id: "chat-run", agentId: "ceo", runtimeMode: "native", status: "succeeded", contextSnapshot: { issueId: "chat" } };
const outputRun = { id: "output-run", agentId: "ceo", runtimeMode: "native", status: "succeeded", contextSnapshot: { issueId: "task" } };
const followUpRun = { id: "follow-up-run", agentId: "ceo", runtimeMode: "native", status: "succeeded", contextSnapshot: { issueId: "chat" } };
const base = { scenarioId: "routines.run-now", origin: "https://fixture.test", companyId: "company", leadId: "ceo", state: { project: { id: "project" } },
tasks: [task], projects: [], goals: [], comments: [], documents: [], revisions: [], humanQueries: [], activity: [], links: ["/ACN/issues/ACN-2"],
routine: { id: "routine" }, invocations: [invocation], outputComments: [{ id: "output", authorAgentId: "ceo", createdByRunId: "output-run", body: "Launch: payment approval is pending." }],
runs: [chatRun, outputRun], reply: "Accepted the Launch summary invocation; its task is waiting.",
toolReads: [
{ runId: "output-run", name: "paperclip_list_projects", at: "2026-01-01T00:01:30.000Z", output: '{"projects":[{"id":"project"}]}' },
{ runId: "follow-up-run", name: "paperclip_get_task", at: "2026-01-01T00:03:00.000Z", output: '{"data":{"id":"task","status":"done"}}' },
] };
const first = structuredClone(base); first.invocations[0]!.status = "issue_created"; first.tasks[0]!.status = "todo";
return { ...base, runs: [...base.runs, followUpRun], state: { ...base.state, replayRunId: "invocation", replayIssueId: "task", firstInvocation: first },
reply: `${first.reply}\nThe summary completed: Launch is still waiting on payment approval. Here is its result task.` };
}
type Observation = ReturnType<typeof valid>;
const withLater = (o: Observation, later: string) => ({ ...o, reply: `${o.state.firstInvocation.reply}\n${later}` });
describe("real manual invocation and delivery distinction", () => {
it("accepts substantive actual result and refuses missing receipts", () => {
expect(gradeIntegratorRoutineInvocation(valid()).passed).toBe(true);
expect(gradeIntegratorRoutineInvocation(undefined).passed).toBe(false);
const o = valid(); o.outputComments = [];
expect(gradeIntegratorRoutineInvocation(o).passed).toBe(false);
const noReads = valid(); delete (noReads as Partial<Observation>).toolReads;
expect(gradeIntegratorRoutineInvocation(noReads).passed).toBe(false);
});
it("does not depend on the wording of a correct later result", () => {
expect(gradeIntegratorRoutineInvocation(withLater(valid(), "RUN-2 is closed and nothing failed. Per its saved comment, the payment sign-off has not come through yet.")).passed).toBe(true);
expect(gradeIntegratorRoutineInvocation(withLater(valid(), "It's done — approval for payment is outstanding.")).passed).toBe(true);
});
it("rejects duplicate delivery, wrong input and unrelated native output", () => {
const duplicate = valid(); duplicate.invocations.push({ ...duplicate.invocations[0]!, id: "another" });
expect(gradeIntegratorRoutineInvocation(duplicate).passed).toBe(false);
const wrongInput = valid(); wrongInput.invocations[0]!.triggerPayload.variables.projectName = "Research";
expect(gradeIntegratorRoutineInvocation(wrongInput).passed).toBe(false);
const wrongReplay = valid(); wrongReplay.state.replayIssueId = "other-task";
expect(gradeIntegratorRoutineInvocation(wrongReplay).passed).toBe(false);
const wrongRun = valid(); wrongRun.runs[1]!.contextSnapshot.issueId = "other-task";
expect(gradeIntegratorRoutineInvocation(wrongRun).passed).toBe(false);
const unread = valid(); unread.toolReads = unread.toolReads.filter(row => row.runId !== "output-run");
expect(gradeIntegratorRoutineInvocation(unread).passed).toBe(false);
});
it("rejects false early completion and a later answer not grounded on the completed result", () => {
const early = valid(); early.state.firstInvocation.reply = "The Launch summary completed successfully.";
early.reply = `${early.state.firstInvocation.reply}\nThe summary completed; payment approval remains pending for Launch.`;
expect(gradeIntegratorRoutineInvocation(early).passed).toBe(false);
const future = valid(); future.state.firstInvocation.reply = "I started the routine; RUN-2 will be marked done once the summary is finished.";
future.reply = `${future.state.firstInvocation.reply}\nThe summary completed; payment approval remains pending.`;
expect(gradeIntegratorRoutineInvocation(future).passed).toBe(true);
const recalled = valid(); recalled.toolReads = recalled.toolReads.filter(row => row.runId !== "follow-up-run");
expect(gradeIntegratorRoutineInvocation(recalled).passed).toBe(false);
const stale = valid(); stale.toolReads[1]!.at = "2026-01-01T00:01:00.000Z";
expect(gradeIntegratorRoutineInvocation(stale).passed).toBe(false);
const noFollowUp = valid(); noFollowUp.runs = noFollowUp.runs.filter(run => run.id !== "follow-up-run");
expect(gradeIntegratorRoutineInvocation(noFollowUp).passed).toBe(false);
const silent = valid(); silent.reply = silent.state.firstInvocation.reply;
expect(gradeIntegratorRoutineInvocation(silent).passed).toBe(false);
const unlinked = valid(); unlinked.links = [];
expect(gradeIntegratorRoutineInvocation(unlinked).passed).toBe(false);
});
});
import { callApiInputDigest } from "./api-read-receipts.js";
describe("call_api reads whose bodies the port redacts", () => {
const apiRead = (runId: string, at: string, input: Record<string, unknown>, status = 200) =>
({ runId, name: "call_api", at, output: "", apiRead: { inputSha256: callApiInputDigest(input), operationId: String(input.operationId), ok: status === 200, status } });
const redacted = (task: Record<string, unknown>, project: Record<string, unknown>, status = 200) => {
const o = valid(); o.toolReads = [apiRead("output-run", "2026-01-01T00:01:30.000Z", project), apiRead("follow-up-run", "2026-01-01T00:03:00.000Z", task, status)] as typeof o.toolReads; return o;
};
const project = { operationId: "GET /api/projects/{id}", pathParams: { id: "project" } };
it("accepts reads of the project and the result task bound by their saved input digests", () => {
expect(gradeIntegratorRoutineInvocation(redacted({ operationId: "GET /api/issues/{id}", pathParams: { id: "ACN-2" } }, project)).passed).toBe(true);
expect(gradeIntegratorRoutineInvocation(redacted({ operationId: "GET /api/issues/{id}", pathParams: { id: "task" } }, project)).passed).toBe(true);
});
it("rejects reads of another record or a failed read", () => {
expect(gradeIntegratorRoutineInvocation(redacted({ operationId: "GET /api/issues/{id}", pathParams: { id: "ACN-1" } }, project)).passed).toBe(false);
expect(gradeIntegratorRoutineInvocation(redacted({ operationId: "GET /api/issues/{id}", pathParams: { id: "task" } }, { ...project, pathParams: { id: "other" } })).passed).toBe(false);
expect(gradeIntegratorRoutineInvocation(redacted({ operationId: "GET /api/issues/{id}", pathParams: { id: "task" } }, project, 404)).passed).toBe(false);
});
});
import { launchSummaryRoutineInput } from "./integrator-routine-invocation.js";
import { createRoutineSchema } from "../../../packages/shared/src/validators/routine.js";
it("validates the real project-name prerequisite as required public text, not unsupported string", () => {
const input = launchSummaryRoutineInput("11111111-1111-4111-8111-111111111111", "22222222-2222-4222-8222-222222222222");
expect(input.variables).toEqual([{ name: "projectName", type: "text", required: true, options: [] }]);
expect(input.assigneeAgentId).toBe("22222222-2222-4222-8222-222222222222");
expect(createRoutineSchema.safeParse({ ...input, variables: [{ name: "projectName", type: "string", required: true }] }).success).toBe(false);
expect(() => launchSummaryRoutineInput("not-a-project-id", input.assigneeAgentId!)).toThrow();
});