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>
99 lines
8.1 KiB
TypeScript
99 lines
8.1 KiB
TypeScript
import { expect } from "@playwright/test";
|
|
import { pollUntil } from "../api.js";
|
|
import { sendChatMessage } from "../chat-flow.js";
|
|
import { checks, type Entity } from "./internal-support.js";
|
|
import { readAgentResource, openStandingChat } from "./identity-support.js";
|
|
import { runtimeDefinition, runtimeObservation, observeRuntime } from "./runtime-support.js";
|
|
|
|
export function gradeReopenConversation(value: unknown) {
|
|
const c = checks(); const o = runtimeObservation(value);
|
|
c.add("persisted-reopen-evidence", !!o, "The existing conversation, original attributed history and navigation result are independently read."); if (!o) return c.result();
|
|
c.add("same-existing-owned-conversation", typeof o.state.existingIssueId === "string" && o.state.reopenedIssueId === o.state.existingIssueId
|
|
&& o.state.reopenedAgentId === o.leadId && o.state.historyRetained === true,
|
|
"The supported result opens the same person's existing Morgan conversation and retains its real prior turn.");
|
|
c.add("navigation-started-no-work", o.state.beforeNavigationRuns === 2 && o.state.afterNavigationRuns === 2
|
|
&& o.state.beforeNavigationChatIds !== undefined && JSON.stringify(o.state.beforeNavigationChatIds) === JSON.stringify(o.state.afterNavigationChatIds)
|
|
&& o.tasks.length === 0,
|
|
"The initial setup and canonical query are the only two runs; navigation creates neither work nor a duplicate standing chat.");
|
|
c.add("private-chat-list-not-readable-to-agent", o.state.agentChatListStatus === 403,
|
|
"The actual agent bearer cannot enumerate the person's private conversation list.");
|
|
// Whether the reply's wording identifies the existing chat is left to a human reader; the
|
|
// grader requires the attributed reply, the opened route and the scope of any chat link.
|
|
c.add("canonical-query-navigation-guidance", typeof o.state.finalQueryReply === "string" && o.state.finalQueryReply.trim().length > 0
|
|
&& o.state.navigationMatchesIntendedAgent === true && o.state.finalNavigationLinksCorrect === true,
|
|
"The canonical query has an attributed reply; the supported human navigation opens Morgan and any provided chat link has the correct scope.");
|
|
return c.result();
|
|
}
|
|
export const runtimeAgentChatCases = [runtimeDefinition("agent-chat.reopen-conversation", async ctx => {
|
|
await ctx.input.api.patch(`/api/agents/${ctx.input.fixtures.agent.id}`, { name: "Morgan" });
|
|
// Genuine history runs before shared pre-canonical funding. Subscription
|
|
// usage is unpriced, so fund both actual scopes before that setup turn.
|
|
for (const scope of [{ scopeType: "company", scopeId: ctx.input.fixtures.company.id },
|
|
{ scopeType: "agent", scopeId: ctx.input.fixtures.agent.id }])
|
|
await ctx.input.api.post(`/api/companies/${ctx.input.fixtures.company.id}/budgets/policies`, {
|
|
...scope, amount: 500, hardStopEnabled: true, isActive: true, unpricedUsagePolicy: "allow" });
|
|
const source = await openStandingChat(ctx);
|
|
const lead = await ctx.input.api.get<Entity>(`/api/agents/${ctx.input.fixtures.agent.id}`);
|
|
const historyMarker = `Retained launch context ${ctx.input.nonce}`;
|
|
await sendChatMessage(ctx.input.page, `Remember this launch reference in our conversation: ${historyMarker}. Acknowledge briefly; no task is needed.`);
|
|
await pollUntil({ label: "Genuine prior Morgan conversation", deadlineAt: Date.now() + 150_000,
|
|
load: async () => { await ctx.refreshIssue(); return { source: ctx.issue(), runs: await ctx.allRuns() }; },
|
|
accept: ({ source, runs }) => source?.conversationState === "waiting" && runs.length === 1 && runs.every(run => run.status === "succeeded") });
|
|
const history = await ctx.input.api.get<Entity[]>(`/api/issues/${source.id}/comments?order=asc`);
|
|
const firstRuns = await ctx.allRuns();
|
|
if (!history.some(row => row.authorAgentId === ctx.input.fixtures.agent.id && firstRuns.some(run => run.id === row.createdByRunId)))
|
|
throw new Error("Existing Morgan history requires a genuine attributed native reply");
|
|
const read = await readAgentResource(ctx, `/api/companies/${ctx.input.fixtures.company.id}/chats`);
|
|
return { existingIssueId: source.id, lead, history, historyMarker, setupRunIds: firstRuns.map(run => run.id), agentChatListStatus: read.status };
|
|
}, gradeReopenConversation, {
|
|
oracleVersion: "paperclip.chat-catalog.runtime.reopen-conversation.v3", authentication: "human-session", setupHumanMessageCount: 1, expectedHumanMessageCount: 2, expectedRunCount: 2, attemptTimeoutMs: 420_000,
|
|
async act(ctx, state) {
|
|
await pollUntil({ label: "Canonical reopen query settled", deadlineAt: Date.now() + 150_000,
|
|
load: async () => { await ctx.refreshIssue(); return { source: ctx.issue(), runs: await ctx.allRuns() }; },
|
|
accept: ({ source, runs }) => source?.conversationState === "waiting" && runs.length === 2 && runs.every(run => run.status === "succeeded") });
|
|
const source = ctx.issue()!; const runs = await ctx.allRuns();
|
|
const comments = await ctx.input.api.get<Entity[]>(`/api/issues/${source.id}/comments?order=asc`);
|
|
const setupIds = state.setupRunIds as string[];
|
|
const finalReplies = comments.filter(row => row.authorAgentId === ctx.input.fixtures.agent.id
|
|
&& runs.some(run => run.id === row.createdByRunId && !setupIds.includes(run.id)));
|
|
state.finalQueryReply = finalReplies.map(row => row.body ?? "").join("\n");
|
|
const allFinalLinks: string[] = [];
|
|
let href: string | undefined;
|
|
for (const reply of finalReplies) {
|
|
const links = await ctx.input.page.locator(`[id=${JSON.stringify(`comment-${reply.id}`)}] a[href]`).evaluateAll(rows => rows.map(row => row.getAttribute("href")!));
|
|
allFinalLinks.push(...links);
|
|
href ??= links.find(link => {
|
|
const url = new URL(link, ctx.input.api.baseURL);
|
|
return url.origin === ctx.input.api.baseURL && url.pathname.split("/").includes("chats")
|
|
&& [ctx.input.fixtures.agent.id, (state.lead as Entity).urlKey].includes(decodeURIComponent(url.pathname.split("/").at(-1)!));
|
|
});
|
|
|
|
}
|
|
const lead = state.lead as Entity;
|
|
state.finalNavigationLinksCorrect = allFinalLinks.filter(link => new URL(link, ctx.input.api.baseURL).pathname.includes("/chats")).every(link => {
|
|
const url = new URL(link, ctx.input.api.baseURL); const parts = url.pathname.split("/").map(decodeURIComponent);
|
|
return url.origin === ctx.input.api.baseURL && parts[1] === ctx.input.fixtures.company.issuePrefix
|
|
&& (parts.at(-1) === "chats" || [lead.id, lead.urlKey].includes(parts.at(-1)));
|
|
});
|
|
state.beforeNavigationRuns = runs.length;
|
|
const chatList = `/api/companies/${ctx.input.fixtures.company.id}/chats`;
|
|
state.beforeNavigationChatIds = (await ctx.input.api.get<Entity[]>(chatList)).map(row => row.id).sort();
|
|
if (href) await ctx.input.page.locator(`[id^="comment-"] a[href=${JSON.stringify(href)}]`).last().click();
|
|
else await ctx.input.page.goto(`/${ctx.input.fixtures.company.issuePrefix}/chats`, { waitUntil: "domcontentloaded" });
|
|
await expect(ctx.input.page.getByTestId("task-chat-composer-input")).toBeVisible();
|
|
await ctx.input.page.reload({ waitUntil: "domcontentloaded" });
|
|
await expect(ctx.input.page.getByTestId("task-chat-composer-input")).toBeVisible();
|
|
const routeRef = decodeURIComponent(new URL(ctx.input.page.url()).pathname.split("/").at(-1)!);
|
|
state.navigationMatchesIntendedAgent = [lead.id, lead.urlKey].includes(routeRef);
|
|
await ctx.refreshIssue(); state.reopenedIssueId = ctx.issue()?.id;
|
|
const reopened = await ctx.input.api.get<{ conversationAgentId: string }>(`/api/issues/${ctx.issue()!.id}`);
|
|
state.reopenedAgentId = reopened.conversationAgentId;
|
|
const after = await ctx.input.api.get<Entity[]>(`/api/issues/${source.id}/comments?order=asc`);
|
|
state.historyRetained = (state.history as Entity[]).every(prior => after.some(row => row.id === prior.id && row.body === prior.body && row.createdByRunId === prior.createdByRunId));
|
|
state.afterNavigationRuns = (await ctx.allRuns()).length;
|
|
state.afterNavigationChatIds = (await ctx.input.api.get<Entity[]>(chatList)).map(row => row.id).sort();
|
|
},
|
|
observe: observeRuntime,
|
|
async verifyResult(ctx) { await expect(ctx.input.page.getByTestId("task-chat-composer-input")).toBeVisible(); },
|
|
})];
|