mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-11-13 15:38:22 +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>
79 lines
6.8 KiB
TypeScript
79 lines
6.8 KiB
TypeScript
import { CHAT_SETUP_HEALTH_ID, CHAT_SETUP_HEALTH_SUITE } from "./chat-setup-health.js";
|
|
import { execFile } from "node:child_process";
|
|
import { promisify } from "node:util";
|
|
import { mkdir, writeFile } from "node:fs/promises";
|
|
import path from "node:path";
|
|
import { reviewedChatProfile } from "./chat-runtime-profile.js";
|
|
import { chatCatalogScenarios } from "./chat-catalog-inventory.js";
|
|
import type { MatrixExecution } from "./types.js";
|
|
|
|
const workflow = "paperclipai/paperclip/.github/workflows/runner-full-stack-e2e.yml@refs/heads/master";
|
|
const testOpsWorkflow = "paperclipai/paperclip/.github/workflows/chat-completion-testops.yml@refs/heads/master";
|
|
const testOpsPilotIds = new Set(["projects.project-work", "projects.project-defaults", "instance-operations.restart-service"]
|
|
.map(scenario => `chat-features.runner-opencode.local.${scenario}`));
|
|
const catalogTestOpsWorkflow = "paperclipai/paperclip/.github/workflows/chat-catalog-testops.yml@refs/heads/master";
|
|
const catalogTestOpsIds = new Set(chatCatalogScenarios.map(row => {
|
|
const task = row.id === "tasks.create-backlog" ? "create-backlog" : row.id === "tasks.delegate-announcement" ? "delegate-announcement" : row.id;
|
|
return `chat-features.runner-opencode.local.${task}`;
|
|
}));
|
|
type Selection = Pick<MatrixExecution, "id" | "suite" | "profile" | "environment" | "task">;
|
|
type Environment = Record<string, string | undefined>;
|
|
export interface RemoteHostDependencies {
|
|
fetch: typeof fetch; platform: string; arch: string;
|
|
sourceRevision: (root: string) => Promise<string>;
|
|
}
|
|
const defaults: RemoteHostDependencies = { fetch: globalThis.fetch, platform: process.platform, arch: process.arch,
|
|
sourceRevision: async root => (await promisify(execFile)("git", ["rev-parse", "HEAD"], { cwd: root, timeout: 5000 })).stdout.trim() };
|
|
/** This new campaign lane is remote-only. Existing profiles/local defaults remain unchanged. */
|
|
export function selectsRemoteChat(executions: readonly Selection[]) {
|
|
return executions.some(e => (e.suite.id === "chat-features" || e.id === CHAT_SETUP_HEALTH_ID) && e.profile.id === "runner-opencode");
|
|
}
|
|
/** Qualifies only the actual trusted GitHub AWS host. The ordinary isolated
|
|
* embedded PostgreSQL bootstrap, migrations and cleanup remain owned by the launcher. */
|
|
export async function prepareChatRemoteHost(input: { executions: readonly Selection[]; env: Environment; repositoryRoot: string; outputDirectory: string; dependencies?: Partial<RemoteHostDependencies> }) {
|
|
if (!selectsRemoteChat(input.executions)) return undefined;
|
|
const deps = { ...defaults, ...input.dependencies }, env = input.env;
|
|
const receipt: Record<string, unknown> = { schema: "paperclip.chat-remote-host/v1", status: "blocked", scope: "before-evaluation", evaluationAdmitted: false,
|
|
allocationSpotOnlyClaimed: false, hostPolicy: "aws-fleet-spot-or-on-demand", databaseLane: "embedded-postgres", automaticRetries: 0, maximumParallel: 2,
|
|
executionIds: input.executions.map(e => e.id) };
|
|
const filename = path.join(input.outputDirectory, "chat-remote-host.json");
|
|
await mkdir(input.outputDirectory, { recursive: true });
|
|
const persist = () => writeFile(filename, `${JSON.stringify(receipt, null, 2)}\n`, { mode: 0o600 });
|
|
let reason = "trusted-github-linux-source-required";
|
|
try {
|
|
const testOpsPilot = (env.GITHUB_WORKFLOW_REF === testOpsWorkflow || env.GITHUB_WORKFLOW_REF === catalogTestOpsWorkflow) && env.PAPERCLIP_CHAT_COMPLETION_TESTOPS === "v1"
|
|
&& env.GITHUB_RUN_ATTEMPT === "1" && /^[1-9][0-9]{0,15}$/.test(env.ALLURE_JOB_RUN_ID ?? "")
|
|
&& Number.isSafeInteger(Number(env.ALLURE_JOB_RUN_ID))
|
|
&& input.executions.every(execution => (env.GITHUB_WORKFLOW_REF === testOpsWorkflow ? testOpsPilotIds : catalogTestOpsIds).has(execution.id)
|
|
&& execution.id === `chat-features.runner-opencode.local.${execution.task.id}`);
|
|
if (env.GITHUB_ACTIONS !== "true" || env.GITHUB_REPOSITORY !== "paperclipai/paperclip" || env.GITHUB_REF !== "refs/heads/master"
|
|
|| env.GITHUB_WORKFLOW_REF !== workflow && !testOpsPilot
|
|
|| !/^\d+$/.test(env.GITHUB_RUN_ID ?? "") || !/^\d+$/.test(env.GITHUB_RUN_ATTEMPT ?? "") || deps.platform !== "linux" || deps.arch !== "x64") throw new Error(reason);
|
|
if (input.executions.some(e => (e.suite.id !== "chat-features" && (e.suite.id !== CHAT_SETUP_HEALTH_SUITE || e.id !== CHAT_SETUP_HEALTH_ID)) || e.profile.id !== "runner-opencode" || !reviewedChatProfile(e.profile) || e.environment.id !== "local" || e.task.automaticRetryPolicy !== "single_attempt")) throw new Error(reason);
|
|
const source = await deps.sourceRevision(input.repositoryRoot);
|
|
if (!/^[a-f0-9]{40}$/.test(source) || env.PAPERCLIP_RUNNER_E2E_SOURCE_SHA !== source) throw new Error(reason);
|
|
Object.assign(receipt, { workflowRef: env.GITHUB_WORKFLOW_REF, workflowRunId: env.GITHUB_RUN_ID, workflowAttempt: Number(env.GITHUB_RUN_ATTEMPT), sourceRevision: source });
|
|
if (testOpsPilot) Object.assign(receipt, { testOps: { admission: "v1", allureJobRunId: env.ALLURE_JOB_RUN_ID } });
|
|
reason = "aws-instance-identity-unavailable";
|
|
const tokenResponse = await deps.fetch("http://169.254.169.254/latest/api/token", { method: "PUT", redirect: "error", headers: { "X-aws-ec2-metadata-token-ttl-seconds": "60" }, signal: AbortSignal.timeout(5000) });
|
|
if (!tokenResponse.ok) throw new Error(reason);
|
|
const token = await tokenResponse.text(); if (!token || token.length > 4096) throw new Error(reason);
|
|
const metadata = async (url: string) => {
|
|
const response = await deps.fetch(url, { redirect: "error", headers: { "X-aws-ec2-metadata-token": token }, signal: AbortSignal.timeout(5000) });
|
|
if (!response.ok) throw new Error(reason); const text = await response.text(); if (text.length > 8192) throw new Error(reason); return text;
|
|
};
|
|
const lifecycle = (await metadata("http://169.254.169.254/latest/meta-data/instance-life-cycle")).trim();
|
|
const identity = JSON.parse(await metadata("http://169.254.169.254/latest/dynamic/instance-identity/document"));
|
|
if (!/^i-[a-f0-9]{8,17}$/.test(identity.instanceId ?? "") || !/^ami-[a-f0-9]{8,17}$/.test(identity.imageId ?? "") || identity.region !== "us-east-1" || identity.architecture !== "x86_64") throw new Error(reason);
|
|
Object.assign(receipt, { instanceId: identity.instanceId, imageId: identity.imageId, region: identity.region, architecture: identity.architecture, lifecycle });
|
|
reason = "unknown-host-lifecycle-refused-before-evaluation";
|
|
if (lifecycle !== "spot" && lifecycle !== "on-demand") throw new Error(reason);
|
|
Object.assign(receipt, { status: "qualified-host-only", spotVerified: lifecycle === "spot" });
|
|
await persist();
|
|
return { receiptPath: filename, databaseLane: "embedded-postgres" as const, maxAutomaticRetries: 0, maxParallel: 2 };
|
|
} catch {
|
|
receipt.reason = reason; await persist();
|
|
throw new Error(`Remote chat preflight blocked: ${reason}; inspect chat-remote-host.json. No evaluation admitted.`);
|
|
}
|
|
}
|