mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-11-15 18:56:05 +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>
63 lines
5.8 KiB
TypeScript
63 lines
5.8 KiB
TypeScript
import { createServer } from "node:http";
|
|
import { createHash, randomBytes } from "node:crypto";
|
|
import { listenOnFetchAllowedPort } from "../../e2e/fetch-allowed-port.js";
|
|
import type { ChatCatalogContext } from "../chat-catalog-contract.js";
|
|
import { root } from "./external-support.js";
|
|
|
|
export interface Tool { name: string; title: string; description: string; readOnly: boolean }
|
|
export interface ProviderCall { method: string; name?: string; authorized: boolean; arguments?: Record<string, unknown>; sequence: number; requestIdSha256?: string; responseHttpStatus?: number; toolError?: boolean; receivedAt?: number; respondedAt?: number; responseSha256?: string }
|
|
export function notesReadResponse(marker: string, unavailable: boolean) {
|
|
return unavailable ? { isError: true, content: [{ type: "text", text: "Synthetic current checklist is unavailable; no current source verification was returned" }] }
|
|
: { content: [{ type: "text", text: marker }] };
|
|
}
|
|
export async function startNotesProvider(marker: string, options: { nonDestructiveWrites?: boolean; readContent?: () => string } = {}) {
|
|
const token = randomBytes(32).toString("hex");
|
|
const calls: ProviderCall[] = []; let writeResult: string | undefined; let failWrites = false; let failReads = false;
|
|
const tools: Tool[] = [
|
|
{ name: "launch_read", title: "Read launch page", description: "Read the synthetic launch page.", readOnly: true },
|
|
{ name: "launch_update", title: "Update launch page", description: "Write the synthetic launch update supplied by the user.", readOnly: false },
|
|
];
|
|
const server = createServer(async (req, res) => {
|
|
try {
|
|
const parts: Buffer[] = []; let length = 0;
|
|
for await (const part of req) { length += part.length; if (length > 65_536) throw new Error("Request too large"); parts.push(part); }
|
|
const message = JSON.parse(Buffer.concat(parts).toString("utf8") || "{}");
|
|
const authorized = req.headers.authorization === `Bearer ${token}`;
|
|
const call: ProviderCall = { method: message.method ?? "unknown", name: message.params?.name, authorized,
|
|
arguments: message.params?.arguments, sequence: calls.length, requestIdSha256: createHash("sha256").update(String(message.id)).digest("hex"), receivedAt: Date.now() }; calls.push(call);
|
|
if (!authorized) { res.writeHead(401); res.end(); return; }
|
|
let result: unknown = {};
|
|
if (message.method === "initialize") result = { protocolVersion: "2024-11-05", capabilities: { tools: {} }, serverInfo: { name: "paperclip-eval-synthetic-notes", version: "1.0.0" } };
|
|
if (message.method === "tools/list") result = { tools: tools.map(tool => ({ name: tool.name, title: tool.title, description: tool.description,
|
|
annotations: { readOnlyHint: tool.readOnly, destructiveHint: !tool.readOnly && !options.nonDestructiveWrites }, inputSchema: { type: "object", properties: { text: { type: "string" } }, additionalProperties: false } })) };
|
|
if (message.method === "tools/call") {
|
|
const tool = tools.find(tool => tool.name === message.params?.name);
|
|
if (!tool) result = { isError: true, content: [{ type: "text", text: "Unknown action" }] };
|
|
else if (tool.readOnly) result = notesReadResponse(options.readContent?.() ?? marker, failReads);
|
|
else if (!tool.readOnly && failWrites) result = { isError: true, content: [{ type: "text", text: "Synthetic write failed; page unchanged" }] };
|
|
else { if (!tool.readOnly) writeResult = message.params?.arguments?.text;
|
|
result = { content: [{ type: "text", text: tool.readOnly ? marker : `Saved synthetic launch update: ${writeResult ?? ""}` }] }; }
|
|
}
|
|
call.responseSha256 = createHash("sha256").update(JSON.stringify(result)).digest("hex"); call.responseHttpStatus = 200; call.toolError = !!(result as { isError?: boolean }).isError; call.respondedAt = Date.now();
|
|
res.writeHead(200, { "content-type": "application/json" }); res.end(JSON.stringify({ jsonrpc: "2.0", id: message.id ?? null, result }));
|
|
} catch { res.writeHead(400); res.end(); }
|
|
});
|
|
const port = await listenOnFetchAllowedPort(server);
|
|
return { url: `http://127.0.0.1:${port}/`, token, tools, calls,
|
|
written: () => writeResult, failReads: () => { failReads = true; }, failWrites: () => { failWrites = true; }, allowWrites: () => { failWrites = false; },
|
|
close: () => new Promise<void>(resolve => server.close(() => resolve())) };
|
|
}
|
|
export type NotesProvider = Awaited<ReturnType<typeof startNotesProvider>>;
|
|
export async function connectNotes(ctx: ChatCatalogContext, provider: NotesProvider, policy: "read-only" | "ask-first" | "all") {
|
|
ctx.input.privateSecrets?.push(provider.token);
|
|
const connection = await ctx.input.api.postSensitive<{ connectionId: string; actions: { readOnly: Array<{ catalogEntryId: string }>; canMakeChanges: Array<{ catalogEntryId: string }> } }>(`${root(ctx)}/tools/apps/connect`, {
|
|
link: provider.url, name: "Synthetic Launch Notes", authMode: "bearer", credentialValues: { "credentials.authorization": provider.token }, grantKind: "organization",
|
|
});
|
|
const read = connection.actions.readOnly.map(row => row.catalogEntryId); const write = connection.actions.canMakeChanges.map(row => row.catalogEntryId);
|
|
if (read.length !== 1 || write.length !== 1) throw new Error("Synthetic provider catalog did not preserve read/write classification");
|
|
await ctx.input.api.post(`${root(ctx)}/tools/apps/${connection.connectionId}/finish`, { enabledCatalogEntryIds: policy === "read-only" ? read : [...read, ...write],
|
|
askFirstCatalogEntryIds: policy === "ask-first" ? write : [], access: { agentIds: [ctx.input.fixtures.agent.id] } });
|
|
await ctx.input.api.putSensitive(`/api/tool-connections/${connection.connectionId}/installs`, { installs: [{ targetType: "agent", targetId: ctx.input.fixtures.agent.id }] });
|
|
return { connectionId: connection.connectionId, read, write };
|
|
}
|