Files
PaperClipAI/tests/runner-e2e/chat-query-attachments.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

99 lines
9.4 KiB
TypeScript

import { readFileSync } from "node:fs";
import { createHash } from "node:crypto";
import { expect, it } from "vitest";
import { attachmentReceiptMatches, chatQueryWithAttachments, resolveChatQueryAttachments, type ChatQueryAttachmentReceipt } from "./chat-query-attachments.js";
import { chatCatalogHumanHistoryMatches, type ChatCatalogContext } from "./chat-catalog-contract.js";
const id = "11111111-1111-4111-8111-111111111111", bytes = Buffer.from("Synthetic website brief");
const receipt: ChatQueryAttachmentReceipt = { id, companyId: "company", issueId: "source", issueCommentId: null,
originalFilename: "website-brief.md", contentPath: `/api/attachments/${id}/content`, byteSize: bytes.length,
sha256: createHash("sha256").update(bytes).digest("hex"), createdByAgentId: null, createdByUserId: "human", originatingRunId: null };
const base = { expected: receipt, saved: receipt, bytes, companyId: "company", issueId: "source" };
it("allows only exact production attachment suffixes after the unchanged canonical query", () => {
const query = "Have Morgan turn the attached brief into a website.";
const body = chatQueryWithAttachments(query, [receipt]);
expect(body).toBe(`${query}\n\n[website-brief.md](${receipt.contentPath})`);
expect(chatQueryWithAttachments(query, [])).toBe(query);
expect(() => chatQueryWithAttachments(query, [receipt, receipt])).toThrow("Only one");
expect(chatQueryWithAttachments(query, [{ ...receipt, originalFilename: "brief[1].md" }])).toContain("[brief\\[1\\].md]");
const history = { implementation: {}, human: [{ id: "comment", body }], runIds: [], boundary: { humanCommentIds: [], runIds: [] }, renderedQuery: body };
expect(chatCatalogHumanHistoryMatches(history)).toBe(true);
expect(chatCatalogHumanHistoryMatches({ ...history, human: [{ id: "comment", body: body + "\nPlease finish" }] })).toBe(false);
expect(chatCatalogHumanHistoryMatches({ ...history, human: [{ id: "comment", body: body.replace(query, "Another query") }] })).toBe(false);
});
it("requires genuine same-source human upload, fetched content hashes and canonical-comment attribution", () => {
expect(attachmentReceiptMatches(base)).toBe(true);
const comment = { id: "canonical", body: "query", authorUserId: "human", authorAgentId: null, createdByRunId: null, issueId: "source", companyId: "company" };
const saved = { ...receipt, issueCommentId: "canonical" };
expect(attachmentReceiptMatches({ ...base, saved, comment })).toBe(true);
for (const patch of [{ companyId: "foreign" }, { issueId: "other-source" }, { contentPath: "https://outside.example/file" },
{ originalFilename: "wrong.md" }, { sha256: "0".repeat(64) }, { createdByAgentId: "agent" }, { originatingRunId: "run" },
{ createdByUserId: "other-human" }, { byteSize: 0 }]) expect(attachmentReceiptMatches({ ...base, saved: { ...receipt, ...patch } })).toBe(false);
expect(attachmentReceiptMatches({ ...base, bytes: Buffer.from("wrong bytes") })).toBe(false);
expect(attachmentReceiptMatches({ ...base, saved: { ...saved, issueCommentId: "other-comment" }, comment })).toBe(false);
expect(attachmentReceiptMatches({ ...base, saved, comment: { ...comment, authorUserId: "other-human" } })).toBe(false);
});
it("independently resolves declared receipt IDs and rejects undeclared, duplicate or forged uploads", async () => {
const ctx = { scenario: { id: "steering.preserve-draft" }, refreshIssue: async () => {}, issue: () => ({ id: "source" }),
input: { fixtures: { company: { id: "company" } }, api: { get: async () => [receipt], getAttachmentContent: async (attachmentId: string) => { expect(attachmentId).toBe(id); return bytes; } } } } as unknown as ChatCatalogContext;
const spec = { stateKey: "brief", filename: "website-brief.md" };
expect(await resolveChatQueryAttachments(ctx, [spec], { brief: receipt })).toEqual([receipt]);
await expect(resolveChatQueryAttachments({ ...ctx, scenario: { ...ctx.scenario, id: "goals.create-goal-tree" } }, [spec], { brief: receipt })).rejects.toThrow("Undeclared");
await expect(resolveChatQueryAttachments(ctx, [spec, spec], { brief: receipt })).rejects.toThrow("duplicate");
await expect(resolveChatQueryAttachments(ctx, [spec], { brief: { ...receipt, sha256: "0".repeat(64) } })).rejects.toThrow("hash");
await expect(resolveChatQueryAttachments(ctx, [spec], {})).rejects.toThrow("persisted UI upload");
});
it.each([
["status-cards.explain-card", "selected-status-card.md"], ["status-cards.archive-card", "selected-status-card.md"],
["routines.webhook-trigger", "selected-routine.md"],
["developer-labs.gated-lab", "historical-lab-link.md"],
["documents-artifacts.resolve-feedback", "selected-source.md"], ["documents-artifacts.restore-document", "selected-source.md"],
["documents-artifacts.produce-download", "selected-source.md"], ["documents-artifacts.revise-delivery", "selected-source.md"],
["skills.author-skill", "skill-context.md"], ["skills.studio-test", "skill-context.md"], ["skills.review-update", "skill-context.md"],
["skills.find-and-install", "skill-context.md"], ["skills.fork-read-only", "skill-context.md"],
["cases.create-case", "customer-bug-brief.md"], ["app-permissions.ask-before-action", "launch-update.md"], ["connection-setup.secure-api-key", "provider-account.md"],
])("binds only the exact factual file for %s", async (scenarioId, filename) => {
const actual = { ...receipt, originalFilename: filename };
const ctx = { scenario: { id: scenarioId }, refreshIssue: async () => {}, issue: () => ({ id: "source" }), input: {
fixtures: { company: { id: "company" } }, api: { get: async () => [actual], getAttachmentContent: async () => bytes },
} } as unknown as ChatCatalogContext;
expect(await resolveChatQueryAttachments(ctx, [{ stateKey: "factualContextAttachment", filename }], { factualContextAttachment: actual })).toEqual([actual]);
await expect(resolveChatQueryAttachments(ctx, [{ stateKey: "factualContextAttachment", filename: "wrong-context.md" }], { factualContextAttachment: actual })).rejects.toThrow("exact scenario filename");
});
it("matches the actual production submittedBody formatter without loading its UI module", () => {
const source = readFileSync(new URL("../../ui/src/components/IssueChatThread.tsx", import.meta.url), "utf8");
const expression = source.match(/const submittedBody = \[\s*trimmed,\s*\.\.\.attachedFiles\.map\(\s*\(item\) =>\s*(`[^`]+`),/);
expect(expression).not.toBeNull();
const production = new Function("trimmed", "attachedFiles", `return [trimmed,...attachedFiles.map(item => ${expression![1]})].filter(Boolean).join("\\n\\n")`) as (query: string, attachments: Array<{ name: string; contentPath: string }>) => string;
for (const name of ["website-brief.md", "brief[1].md", "brief (draft).md"]) {
expect(chatQueryWithAttachments("Original query", [{ ...receipt, originalFilename: name }])).toBe(production("Original query", [{ name, contentPath: receipt.contentPath }]));
}
});
it("also matches the active TaskChatComposer for its one declared upload", () => {
const source = readFileSync(new URL("../../ui/src/components/task-chat/TaskChatComposer.tsx", import.meta.url), "utf8");
const escapeBody = source.match(/function escapeMarkdownLabel\(name: string\): string \{\s*([^}]+)\}/)![1]!;
const escape = new Function("name", escapeBody);
const template = source.match(/const refLines = attachedRefs\s*\.map\(\(item\) => (`[^`]+`)\)/)![1]!;
const production = new Function("trimmed", "attachedRefs", "escapeMarkdownLabel", "isImage", `return [trimmed,attachedRefs.map(item => ${template}).join("\\n")].filter(Boolean).join("\\n\\n")`) as (query: string, refs: Array<{ name: string; contentPath: string }>, escape: Function, isImage: Function) => string;
// The composer marks images with `!`; the declared uploads are Markdown documents.
expect(source).toContain("function isImage(item: ComposerAttachment)");
const isImage = (item: { name: string }) => /\.(png|jpe?g|gif|webp|heic|heif)$/i.test(item.name);
for (const name of ["website-brief.md", "brief[1].md"]) expect(chatQueryWithAttachments("Canonical query", [{ ...receipt, originalFilename: name }])).toBe(production("Canonical query", [{ name, contentPath: receipt.contentPath }], escape, isImage));
});
it("admits only the declared human company-preview package with independent receipt checks", async () => {
const actual = { ...receipt, originalFilename: "company-package.json" };
const ctx = { scenario: { id: "companies.import-preview" }, refreshIssue: async () => {}, issue: () => ({ id: "source" }), input: {
fixtures: { company: { id: "company" } }, api: { get: async () => [actual], getAttachmentContent: async () => bytes },
} } as unknown as ChatCatalogContext;
const spec = { stateKey: "packageAttachment", filename: "company-package.json" };
expect(await resolveChatQueryAttachments(ctx, [spec], { packageAttachment: actual })).toEqual([actual]);
for (const invalid of [{ ...spec, filename: "arbitrary.zip" }, { ...spec, stateKey: "otherUpload" }])
await expect(resolveChatQueryAttachments(ctx, [invalid], { packageAttachment: actual })).rejects.toThrow("exact declared");
await expect(resolveChatQueryAttachments(ctx, [spec, spec], { packageAttachment: actual })).rejects.toThrow("duplicate");
await expect(resolveChatQueryAttachments(ctx, [spec], { packageAttachment: { ...actual, createdByUserId: "forged" } })).rejects.toThrow("attribution");
await expect(resolveChatQueryAttachments(ctx, [spec], { packageAttachment: { ...actual, sha256: "0".repeat(64) } })).rejects.toThrow("hash");
});