mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-11-13 15:55:46 +01:00
## 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>
73 lines
7.6 KiB
TypeScript
73 lines
7.6 KiB
TypeScript
import { expect } from "@playwright/test";
|
|
import type { ChatCatalogContext } from "../chat-catalog-contract.js";
|
|
import type { PluginRecord } from "../../../packages/shared/src/types/plugin.js";
|
|
import { pollUntil } from "../api.js";
|
|
import { definition, externalObservation, observeExternal, root, checks, verifyRenderedReply } from "./external-support.js";
|
|
import { createPreviewFixture } from "./external-plugin-fixture.js";
|
|
import { probeAgentPermission, waitForReply } from "./identity-support.js";
|
|
const fixtures = new WeakMap<ChatCatalogContext, Awaited<ReturnType<typeof createPreviewFixture>>>();
|
|
export function gradePluginInstall(value: unknown) {
|
|
const c = checks(), o = externalObservation(value);
|
|
c.add("complete-plugin-install-evidence", !!o, "Vetted package, administrator admission, real activation and visible installed state are required."); if (!o) return c.result();
|
|
const plugin = o.state.pluginAfter as PluginRecord | undefined;
|
|
c.add("exact-vetted-package-installed", o.state.syntheticPackage === true && plugin?.id === o.state.pluginId && plugin?.pluginKey === o.state.pluginKey
|
|
&& plugin?.status === "ready" && plugin?.manifestJson?.displayName === "Website Preview Eval", "The exact CLI-scaffolded package is loaded and activated by the real product loader.");
|
|
c.add("administrator-installation-only", o.state.operatorChannel === "board-api-local-path" && o.state.agentInstallStatus === 403 && o.state.absentBeforeHuman === true && o.state.humanInstallStatus === 200
|
|
&& !!plugin && o.activity.some(row => row.entityId === plugin.id && row.actorType === "user" && row.action === "plugin.installed"),
|
|
"The ordinary agent cannot install the package; the authenticated operator management request has independent mutation attribution.");
|
|
c.add("actual-contribution-and-visible-state", o.state.previewReadback === true && o.state.humanSawActivation === true,
|
|
"The human sees the installed activation state and the real company-scoped contributed route answers its fixture marker.");
|
|
// Graded from saved state and setup-chosen literals, never reply wording: the
|
|
// package was still absent when chat answered (administrator-installation-only).
|
|
const literals = [o.state.packageName, o.state.displayName, o.state.pluginKey, o.state.packageRoot].filter((value): value is string => typeof value === "string" && value.length > 0);
|
|
c.add("truthful-installation-handoff", literals.length === 4 && o.phaseReplies.some(row => literals.some(literal => row.body.includes(literal))),
|
|
"Chat answered before the human installation step and named the exact package (its name, key or local path).");
|
|
return c.result();
|
|
}
|
|
export const pluginInstallExternalCases = [definition("plugins.install-plugin", async ctx => {
|
|
const fixture = await createPreviewFixture(ctx); fixtures.set(ctx, fixture);
|
|
try {
|
|
const installed = await ctx.input.api.get<PluginRecord[]>("/api/plugins");
|
|
if (installed.some(row => row.pluginKey === fixture.key)) throw new Error("Disposable plugin already installed before its installation journey");
|
|
const agentInstallStatus = await probeAgentPermission(ctx, "/api/plugins/install", "POST", { packageName: fixture.packageRoot, isLocalPath: true });
|
|
await ctx.input.api.post(`${root(ctx)}/chats/${ctx.input.fixtures.agent.id}`); await ctx.refreshIssue();
|
|
await ctx.input.api.patch(`/api/issues/${ctx.issue()!.id}`, { description: `The selected vetted synthetic package is Website Preview Eval (@paperclip-eval/website-preview), key ${fixture.key}, available as a local package at ${fixture.packageRoot}. It is not installed. Local-path installation uses the administrator plugin-management API. No external package needs downloading.` });
|
|
return { pluginKey: fixture.key, packageName: "@paperclip-eval/website-preview", displayName: "Website Preview Eval", packageRoot: fixture.packageRoot, syntheticPackage: true, agentInstallStatus };
|
|
} catch (error) { fixtures.delete(ctx); await fixture.close(); throw error; }
|
|
}, gradePluginInstall, {
|
|
oracleVersion: "paperclip.chat-catalog.external-plugin-install.v2",
|
|
sourceFiles: ["chat-catalog-domains/external-plugin-install.ts", "chat-catalog-domains/external-plugin-fixture.ts", "chat-catalog-domains/external-support.ts", "chat-catalog-domains/internal-support.ts", "chat-catalog-domains/identity-support.ts", "../../cli/node_modules/tsx/dist/loader.mjs", "../../cli/src/index.ts", "../../packages/plugins/create-paperclip-plugin/src/index.ts", "../../packages/plugins/sdk/src/index.ts", "../../packages/plugins/sdk/dist/index.js", "chat-scenarios/plugins.yaml"],
|
|
async act(ctx, state) {
|
|
await waitForReply(ctx);
|
|
const fixture = fixtures.get(ctx); if (!fixture) throw new Error("Vetted package fixture is missing");
|
|
const installed = await ctx.input.api.get<PluginRecord[]>("/api/plugins"); state.absentBeforeHuman = !installed.some(row => row.pluginKey === fixture.key);
|
|
if (!state.absentBeforeHuman) throw new Error("Plugin installed before the authorized human management step");
|
|
// The visible npm dialog does not accept localPath; the public board API is
|
|
// the supported local-package installation path, under the same human session.
|
|
const response = await ctx.input.api.request.post("/api/plugins/install", { data: { packageName: fixture.packageRoot, isLocalPath: true } });
|
|
state.operatorChannel = "board-api-local-path"; state.humanInstallStatus = response.status(); if (!response.ok()) throw new Error("Operator package installation failed; error details withheld");
|
|
const plugin = await response.json() as PluginRecord; state.pluginId = plugin.id;
|
|
await pollUntil({ label: "operator-installed synthetic plugin activation", deadlineAt: Date.now() + 45_000,
|
|
load: () => ctx.input.api.get<PluginRecord>(`/api/plugins/${plugin.id}`), accept: row => row.status === "ready", reject: row => row.status === "error" ? "Synthetic plugin activation failed; inspect public plugin status" : undefined });
|
|
const source = ctx.input.page.url(); await ctx.input.page.goto(`/${ctx.input.fixtures.company.issuePrefix}/company/settings/instance/plugins`, { waitUntil: "domcontentloaded" });
|
|
const row = ctx.input.page.locator("li").filter({ hasText: "Website Preview Eval" });
|
|
await expect(row.getByRole("link", { name: "Website Preview Eval", exact: true })).toBeVisible(); await expect(row.getByRole("button", { name: "Disable", exact: true })).toBeVisible(); state.humanSawActivation = true;
|
|
await ctx.input.evidence("chat-catalog-preview-plugin-install.json", { operatorChannel: state.operatorChannel, pluginId: plugin.id, pluginKey: fixture.key, status: "ready", visibleManagerState: "Disable" });
|
|
await ctx.input.page.goto(source, { waitUntil: "domcontentloaded" });
|
|
},
|
|
async observe(ctx, state) {
|
|
const fixture = fixtures.get(ctx); if (!fixture) throw new Error("Vetted package fixture is missing");
|
|
state.pluginAfter = await ctx.input.api.get<PluginRecord>(`/api/plugins/${state.pluginId}`);
|
|
const route = await ctx.input.api.request.get(`/api/plugins/${state.pluginId}/api/preview?companyId=${ctx.input.fixtures.company.id}`);
|
|
const result = route.ok() ? await route.json() as { companyId?: string; marker?: string } : undefined;
|
|
state.previewReadback = route.ok() && result?.companyId === ctx.input.fixtures.company.id && result?.marker === fixture.marker;
|
|
return observeExternal(ctx, state);
|
|
},
|
|
async verifyResult(ctx, _state, value) { await verifyRenderedReply(ctx, value); },
|
|
async cleanup(ctx, state) {
|
|
const fixture = fixtures.get(ctx); fixtures.delete(ctx);
|
|
try { if (typeof state.pluginId === "string") await ctx.input.api.delete(`/api/plugins/${state.pluginId}`, { allowNotFound: true }); }
|
|
finally { await fixture?.close(); }
|
|
},
|
|
})];
|