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

47 lines
7.2 KiB
TypeScript

import { expect, it } from "vitest";
import { deliveryRevisionBound, gradeInterruption } from "./runtime-interruption.js";
function valid() {
return { scenarioId: "steering.interrupt-with-direction", companyId: "company", leadId: "ceo", agents: [], links: [], humanQueries: [], reply: "", tasks: [{ id: "website", status: "done" }], activity: [],
runs: [{ id: "old", status: "cancelled" }, { id: "source", status: "succeeded" }, { id: "new", status: "succeeded" }],
state: { task: { id: "website" }, worker: { id: "morgan" }, workerRunId: "old", approvedDirection: { id: "accepted-plan" },
directionQueue: { issueId: "website", targetRunId: "old", protocol: "paperclip_runner_v1", entries: [{ source: { kind: "interaction", requiresFreshSession: true, interactionId: "accepted-plan" } }] },
planApprovedByHuman: true, interruptControlEnabled: true, humanInterrupted: true, interruptStatus: 200,
oldRunAfter: { id: "old", status: "cancelled", finishedAt: "2026-10-26T12:00:00.000Z" }, nativeStopAcknowledged: true, oldEvidenceRetained: true,
queryReply: "Use the supported Interrupt button beside your saved simpler plan response. It stops the current attempt before sending the accepted direction to one eligible fresh continuation.", workerRunningAfterReply: true, humanFollowedReplyTaskLink: true,
successorCount: 1, successor: { id: "new", agentId: "morgan", status: "succeeded", startedAt: "2026-10-26T12:00:01.000Z" }, neverConcurrentOwners: true,
acceptedDirectionInputBound: true, freshSessionBound: true, queueEmptyAfterDelivery: true, actualSimplePage: true, planRevisionRetained: true, deliveryRevisionBound: true } };
}
it("accepts a genuine fresh-response Interrupt and one post-cleanup owner", () => expect(gradeInterruption(valid()).passed).toBe(true));
it("accepts natural human control guidance", () => { const o = valid(); o.state.queryReply = "You can stop this execution with Interrupt on the waiting approved plan; after cleanup the next turn receives the saved simpler layout."; expect(gradeInterruption(o).passed).toBe(true); });
it("rejects a legacy protocol substitute", () => { const o = valid(); o.state.directionQueue.protocol = "legacy"; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects an invented fresh-session response", () => { const o = valid(); o.state.directionQueue.entries[0]!.source.requiresFreshSession = false; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects pending direction without actual human approval", () => { const o = valid(); o.state.planApprovedByHuman = false; expect(gradeInterruption(o).passed).toBe(false); });
it("does not grade the reply's wording", () => { const o = valid(); o.state.queryReply = "Open [the task](/COM/issues/COM-1) and press the control beside the queued plan."; expect(gradeInterruption(o).passed).toBe(true); });
it("rejects the agent stopping the run itself before the person acts", () => { const o = valid(); o.state.workerRunningAfterReply = false; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects a reply that leaves the person no rendered link to the task's Interrupt control", () => { const o = valid(); o.state.humanFollowedReplyTaskLink = false; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects a stop request without confirmed native acknowledgement", () => { const o = valid(); o.state.nativeStopAcknowledged = false; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects starting the new owner before the original settles", () => { const o = valid(); o.state.successor.startedAt = "2026-10-26T11:59:59.000Z"; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects observed concurrent ownership", () => { const o = valid(); o.state.neverConcurrentOwners = false; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects duplicate accepted-direction consumers", () => { const o = valid(); o.state.successorCount = 2; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects consuming a different plan revision", () => { const o = valid(); o.state.acceptedDirectionInputBound = false; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects continuing the old session despite the fresh bound response", () => { const o = valid(); o.state.freshSessionBound = false; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects rewriting the accepted plan after the interruption", () => { const o = valid(); o.state.planRevisionRetained = false; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects a friendly completion with no real simpler page", () => { const o = valid(); o.state.actualSimplePage = false; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects lost original run history", () => { const o = valid(); o.state.oldEvidenceRetained = false; expect(gradeInterruption(o).passed).toBe(false); });
it("rejects missing actual interruption evidence", () => expect(gradeInterruption(undefined).passed).toBe(false));
// Shaped like run 42592781763: the successor saved website-delivery in its own
// words, without the plan's setup marker.
function delivery() {
return { documents: [{ id: "plan-doc", key: "plan", issueId: "website", latestRevisionId: "plan-r1" }, { id: "brief-doc", key: "website-brief", issueId: "website", latestRevisionId: "brief-r1" }, { id: "delivery-doc", key: "website-delivery", issueId: "website", latestRevisionId: "delivery-r1" }],
revisions: [{ id: "plan-r1", documentId: "plan-doc", createdByRunId: null, createdByAgentId: null, body: "# Approved simpler website direction\nSimpler layout nonce-1" },
{ id: "brief-r1", documentId: "brief-doc", createdByRunId: "new", createdByAgentId: "morgan", body: "# Website brief" },
{ id: "delivery-r1", documentId: "delivery-doc", createdByRunId: "new", createdByAgentId: "morgan", body: "# Website delivery\n- Deliverable: website.html, single semantic page." }] };
}
it("binds the successor's own current website-delivery revision without a copied setup marker", () => expect(deliveryRevisionBound(delivery(), "website", "new", "morgan")).toBe(true));
it("accepts reworded delivery prose", () => { const o = delivery(); (o.revisions[2] as { body: string }).body = "Shipped the one-page Acorn Notes site."; expect(deliveryRevisionBound(o, "website", "new", "morgan")).toBe(true); });
it("rejects a delivery document written by the stopped attempt", () => { const o = delivery(); o.revisions[2]!.createdByRunId = "old"; expect(deliveryRevisionBound(o, "website", "new", "morgan")).toBe(false); });
it("rejects a delivery written by another agent", () => { const o = delivery(); o.revisions[2]!.createdByAgentId = "ceo"; expect(deliveryRevisionBound(o, "website", "new", "morgan")).toBe(false); });
it("rejects a superseded successor revision", () => { const o = delivery(); o.documents[2]!.latestRevisionId = "later"; expect(deliveryRevisionBound(o, "website", "new", "morgan")).toBe(false); });
it("rejects only some other successor document", () => { const o = delivery(); o.documents.pop(); expect(deliveryRevisionBound(o, "website", "new", "morgan")).toBe(false); });
it("rejects a delivery document on a different task", () => { const o = delivery(); o.documents[2]!.issueId = "other"; expect(deliveryRevisionBound(o, "website", "new", "morgan")).toBe(false); });