sandbox-agent/foundry/packages/client/test/e2e/workbench-load-e2e.test.ts
Nathan Flurry ae191d1ae1
Refactor Foundry GitHub state and sandbox runtime (#247)
* Move Foundry HTTP APIs out of /api/rivet

* Move Foundry HTTP APIs onto /v1

* Fix Foundry Rivet base path and frontend endpoint fallback

* Configure Foundry Rivet runner pool for /v1

* Remove Foundry Rivet runner override

* Serve Foundry Rivet routes directly from Bun

* Log Foundry RivetKit deployment friction

* Add actor display metadata

* Tighten actor schema constraints

* Reset actor persistence baseline

* Remove temporary actor key version prefix

Railway has no persistent volumes so stale actors are wiped on
each deploy. The v2 key rotation is no longer needed.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Cache app workspace actor handle across requests

Every request was calling getOrCreate on the Rivet engine API
to resolve the workspace actor, even though it's always the same
actor. Cache the handle and invalidate on error so retries
re-resolve. This eliminates redundant cross-region round-trips
to api.rivet.dev on every request.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add temporary debug logging to GitHub OAuth exchange

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Make squashed baseline migrations idempotent

Use CREATE TABLE IF NOT EXISTS and CREATE UNIQUE INDEX IF NOT
EXISTS so the squashed baseline can run against actors that
already have tables from the pre-squash migration sequence.
This fixes the "table already exists" error when org workspace
actors wake up with stale migration journals.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Revert "Make squashed baseline migrations idempotent"

This reverts commit 356c146035.

* Fix GitHub OAuth callback by removing retry wrapper

OAuth authorization codes are single-use. The appWorkspaceAction wrapper
retries failed calls up to 20 times, but if the code exchange succeeds
and a later step fails, every retry sends the already-consumed code,
producing "bad_verification_code" from GitHub.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add runner versioning to RivetKit registry

Uses Date.now() so each process start gets a unique version.
This ensures Rivet Cloud migrates actors to the new runner on
deploy instead of routing requests to stale runners.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add backend request and workspace logging

* Log callback request headers

* Make GitHub OAuth callback idempotent against duplicate requests

Clear oauthState before exchangeCode so duplicate callback requests
fail the state check instead of hitting GitHub with a consumed code.
Marked as HACK — root cause of duplicate HTTP requests is unknown.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add temporary header dump on GitHub OAuth callback

Log all request headers on the callback endpoint to diagnose
the source of duplicate requests (Railway proxy, Cloudflare, browser).
Remove once root cause is identified.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Defer slow GitHub org sync to workflow queue for fast OAuth callback

Split syncGithubSessionFromToken into a fast path (initGithubSession:
exchange code, get viewer, store token+identity) and a slow path
(syncGithubOrganizations: list orgs/installations, sync workspaces).

completeAppGithubAuth now returns the 302 redirect in ~2s instead of
~18s by enqueuing the org sync to the workspace workflow queue
(fire-and-forget). This eliminates the proxy timeout window that was
causing duplicate callback requests.

bootstrapAppGithubSession (dev-only) still calls the full synchronous
sync since proxy timeouts are not a concern and it needs the session
fully populated before returning.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* foundry: async app repo import on org select

* foundry: parallelize app snapshot org reads

* repo: push all current workspace changes

* foundry: update runner version and snapshot logging

* Refactor Foundry GitHub state and sandbox runtime

Refactors Foundry around organization/repository ownership and adds an organization-scoped GitHub state actor plus a user-scoped GitHub auth actor, removing the old project PR/branch sync actors and repo PR cache.

Updates sandbox provisioning to rely on sandbox-agent for in-sandbox work, hardens Daytona startup and image-build behavior, and surfaces runtime and task-startup errors more clearly in the UI.

Extends workbench and GitHub state handling to track merged PR state, adds runtime-issue tracking, refreshes client/test/config wiring, and documents the main live Foundry test flow plus actor coordination rules.

Also updates the remaining Sandbox Agent install-version references in docs/examples to the current pinned minor channel.

Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-13 02:45:07 -07:00

319 lines
11 KiB
TypeScript

import { describe, expect, it } from "vitest";
import {
createFoundryLogger,
type TaskWorkbenchSnapshot,
type WorkbenchAgentTab,
type WorkbenchTask,
type WorkbenchModelId,
type WorkbenchTranscriptEvent,
} from "@sandbox-agent/foundry-shared";
import { createBackendClient } from "../../src/backend-client.js";
const RUN_WORKBENCH_LOAD_E2E = process.env.HF_ENABLE_DAEMON_WORKBENCH_LOAD_E2E === "1";
const logger = createFoundryLogger({
service: "foundry-client-e2e",
bindings: {
suite: "workbench-load",
},
});
function requiredEnv(name: string): string {
const value = process.env[name]?.trim();
if (!value) {
throw new Error(`Missing required env var: ${name}`);
}
return value;
}
function workbenchModelEnv(name: string, fallback: WorkbenchModelId): WorkbenchModelId {
const value = process.env[name]?.trim();
switch (value) {
case "claude-sonnet-4":
case "claude-opus-4":
case "gpt-4o":
case "o3":
return value;
default:
return fallback;
}
}
function intEnv(name: string, fallback: number): number {
const raw = process.env[name]?.trim();
if (!raw) {
return fallback;
}
const value = Number.parseInt(raw, 10);
return Number.isFinite(value) && value > 0 ? value : fallback;
}
async function sleep(ms: number): Promise<void> {
await new Promise((resolve) => setTimeout(resolve, ms));
}
async function poll<T>(label: string, timeoutMs: number, intervalMs: number, fn: () => Promise<T>, isDone: (value: T) => boolean): Promise<T> {
const startedAt = Date.now();
let lastValue: T;
for (;;) {
lastValue = await fn();
if (isDone(lastValue)) {
return lastValue;
}
if (Date.now() - startedAt > timeoutMs) {
throw new Error(`timed out waiting for ${label}`);
}
await sleep(intervalMs);
}
}
function findTask(snapshot: TaskWorkbenchSnapshot, taskId: string): WorkbenchTask {
const task = snapshot.tasks.find((candidate) => candidate.id === taskId);
if (!task) {
throw new Error(`task ${taskId} missing from snapshot`);
}
return task;
}
function findTab(task: WorkbenchTask, tabId: string): WorkbenchAgentTab {
const tab = task.tabs.find((candidate) => candidate.id === tabId);
if (!tab) {
throw new Error(`tab ${tabId} missing from task ${task.id}`);
}
return tab;
}
function extractEventText(event: WorkbenchTranscriptEvent): string {
const payload = event.payload;
if (!payload || typeof payload !== "object") {
return String(payload ?? "");
}
const envelope = payload as {
method?: unknown;
params?: unknown;
result?: unknown;
};
const params = envelope.params;
if (params && typeof params === "object") {
const update = (params as { update?: unknown }).update;
if (update && typeof update === "object") {
const content = (update as { content?: unknown }).content;
if (content && typeof content === "object") {
const chunkText = (content as { text?: unknown }).text;
if (typeof chunkText === "string") {
return chunkText;
}
}
}
const text = (params as { text?: unknown }).text;
if (typeof text === "string" && text.trim()) {
return text.trim();
}
const prompt = (params as { prompt?: Array<{ text?: unknown }> }).prompt;
if (Array.isArray(prompt)) {
return prompt
.map((item) => (typeof item?.text === "string" ? item.text.trim() : ""))
.filter(Boolean)
.join("\n");
}
}
const result = envelope.result;
if (result && typeof result === "object") {
const text = (result as { text?: unknown }).text;
if (typeof text === "string" && text.trim()) {
return text.trim();
}
}
return typeof envelope.method === "string" ? envelope.method : JSON.stringify(payload);
}
function transcriptIncludesAgentText(transcript: WorkbenchTranscriptEvent[], expectedText: string): boolean {
return transcript
.filter((event) => event.sender === "agent")
.map((event) => extractEventText(event))
.join("")
.includes(expectedText);
}
function average(values: number[]): number {
return values.reduce((sum, value) => sum + value, 0) / Math.max(values.length, 1);
}
async function measureWorkbenchSnapshot(
client: ReturnType<typeof createBackendClient>,
workspaceId: string,
iterations: number,
): Promise<{
avgMs: number;
maxMs: number;
payloadBytes: number;
taskCount: number;
tabCount: number;
transcriptEventCount: number;
}> {
const durations: number[] = [];
let snapshot: TaskWorkbenchSnapshot | null = null;
for (let index = 0; index < iterations; index += 1) {
const startedAt = performance.now();
snapshot = await client.getWorkbench(workspaceId);
durations.push(performance.now() - startedAt);
}
const finalSnapshot = snapshot ?? {
workspaceId,
repos: [],
projects: [],
tasks: [],
};
const payloadBytes = Buffer.byteLength(JSON.stringify(finalSnapshot), "utf8");
const tabCount = finalSnapshot.tasks.reduce((sum, task) => sum + task.tabs.length, 0);
const transcriptEventCount = finalSnapshot.tasks.reduce((sum, task) => sum + task.tabs.reduce((tabSum, tab) => tabSum + tab.transcript.length, 0), 0);
return {
avgMs: Math.round(average(durations)),
maxMs: Math.round(Math.max(...durations, 0)),
payloadBytes,
taskCount: finalSnapshot.tasks.length,
tabCount,
transcriptEventCount,
};
}
describe("e2e(client): workbench load", () => {
it.skipIf(!RUN_WORKBENCH_LOAD_E2E)("runs a simple sequential load profile against the real backend", { timeout: 30 * 60_000 }, async () => {
const endpoint = process.env.HF_E2E_BACKEND_ENDPOINT?.trim() || "http://127.0.0.1:7741/v1/rivet";
const workspaceId = process.env.HF_E2E_WORKSPACE?.trim() || "default";
const repoRemote = requiredEnv("HF_E2E_GITHUB_REPO");
const model = workbenchModelEnv("HF_E2E_MODEL", "gpt-4o");
const taskCount = intEnv("HF_LOAD_TASK_COUNT", 3);
const extraSessionCount = intEnv("HF_LOAD_EXTRA_SESSION_COUNT", 2);
const pollIntervalMs = intEnv("HF_LOAD_POLL_INTERVAL_MS", 2_000);
const client = createBackendClient({
endpoint,
defaultWorkspaceId: workspaceId,
});
const repo = await client.addRepo(workspaceId, repoRemote);
const createTaskLatencies: number[] = [];
const provisionLatencies: number[] = [];
const createSessionLatencies: number[] = [];
const messageRoundTripLatencies: number[] = [];
const snapshotSeries: Array<{
taskCount: number;
avgMs: number;
maxMs: number;
payloadBytes: number;
tabCount: number;
transcriptEventCount: number;
}> = [];
snapshotSeries.push(await measureWorkbenchSnapshot(client, workspaceId, 2));
for (let taskIndex = 0; taskIndex < taskCount; taskIndex += 1) {
const runId = `load-${taskIndex}-${Date.now().toString(36)}`;
const initialReply = `LOAD_INIT_${runId}`;
const createStartedAt = performance.now();
const created = await client.createWorkbenchTask(workspaceId, {
repoId: repo.repoId,
title: `Workbench Load ${runId}`,
branch: `load/${runId}`,
model,
task: `Reply with exactly: ${initialReply}`,
});
createTaskLatencies.push(performance.now() - createStartedAt);
const provisionStartedAt = performance.now();
const provisioned = await poll(
`task ${runId} provisioning`,
12 * 60_000,
pollIntervalMs,
async () => findTask(await client.getWorkbench(workspaceId), created.taskId),
(task) => {
const tab = task.tabs[0];
return Boolean(tab && task.status === "idle" && tab.status === "idle" && transcriptIncludesAgentText(tab.transcript, initialReply));
},
);
provisionLatencies.push(performance.now() - provisionStartedAt);
expect(provisioned.tabs.length).toBeGreaterThan(0);
const primaryTab = provisioned.tabs[0]!;
expect(transcriptIncludesAgentText(primaryTab.transcript, initialReply)).toBe(true);
for (let sessionIndex = 0; sessionIndex < extraSessionCount; sessionIndex += 1) {
const expectedReply = `LOAD_REPLY_${runId}_${sessionIndex}`;
const createSessionStartedAt = performance.now();
const createdSession = await client.createWorkbenchSession(workspaceId, {
taskId: created.taskId,
model,
});
createSessionLatencies.push(performance.now() - createSessionStartedAt);
await client.sendWorkbenchMessage(workspaceId, {
taskId: created.taskId,
tabId: createdSession.tabId,
text: `Run pwd in the repo, then reply with exactly: ${expectedReply}`,
attachments: [],
});
const messageStartedAt = performance.now();
const withReply = await poll(
`task ${runId} session ${sessionIndex} reply`,
10 * 60_000,
pollIntervalMs,
async () => findTask(await client.getWorkbench(workspaceId), created.taskId),
(task) => {
const tab = findTab(task, createdSession.tabId);
return tab.status === "idle" && transcriptIncludesAgentText(tab.transcript, expectedReply);
},
);
messageRoundTripLatencies.push(performance.now() - messageStartedAt);
expect(transcriptIncludesAgentText(findTab(withReply, createdSession.tabId).transcript, expectedReply)).toBe(true);
}
const snapshotMetrics = await measureWorkbenchSnapshot(client, workspaceId, 3);
snapshotSeries.push(snapshotMetrics);
logger.info(
{
taskIndex: taskIndex + 1,
...snapshotMetrics,
},
"workbench_load_snapshot",
);
}
const firstSnapshot = snapshotSeries[0]!;
const lastSnapshot = snapshotSeries[snapshotSeries.length - 1]!;
const summary = {
taskCount,
extraSessionCount,
createTaskAvgMs: Math.round(average(createTaskLatencies)),
provisionAvgMs: Math.round(average(provisionLatencies)),
createSessionAvgMs: Math.round(average(createSessionLatencies)),
messageRoundTripAvgMs: Math.round(average(messageRoundTripLatencies)),
snapshotReadBaselineAvgMs: firstSnapshot.avgMs,
snapshotReadFinalAvgMs: lastSnapshot.avgMs,
snapshotReadFinalMaxMs: lastSnapshot.maxMs,
snapshotPayloadBaselineBytes: firstSnapshot.payloadBytes,
snapshotPayloadFinalBytes: lastSnapshot.payloadBytes,
snapshotTabFinalCount: lastSnapshot.tabCount,
snapshotTranscriptFinalCount: lastSnapshot.transcriptEventCount,
};
logger.info(summary, "workbench_load_summary");
expect(createTaskLatencies.length).toBe(taskCount);
expect(provisionLatencies.length).toBe(taskCount);
expect(createSessionLatencies.length).toBe(taskCount * extraSessionCount);
expect(messageRoundTripLatencies.length).toBe(taskCount * extraSessionCount);
});
});