Files
PaperClipAI/ui/src/api/voiceSessions.test.ts
T
DottaandPaperclip 22d05644f4 feat(ui): add Speko voice setup, task calls, and Storybook coverage (#15728)
Runtime #15727 is merged. This PR adds the UI. Qualification follows in
#15729.

## Thinking Path

> - Paperclip manages agent work through tasks and governed connections.
> - Chat connections already bind external conversations to those tasks.
> - Speko handles voice and telephony, while the selected Paperclip
agent does the work.
> - A call needs durable session authority, ordered follow-ups, and safe
result delivery.
> - This PR adds the Speko setup and call controls to the existing Apps
and task UI.
> - This PR depends on #15727. The qualification suite follows
separately. Speko stays experimental.

## Linked Issues or Issue Description

**Problem or motivation**

Phone and browser voice conversations cannot use the chat connection
task lifecycle. A separate voice runtime is needed to preserve task
authority, retries, and reply history.

**Proposed solution**

Add company-scoped voice sessions, signed callbacks, durable replies,
inbound intake, and an agent-facing outbound-call tool. Reuse existing
tasks, queues, secrets, and publications.

**Roadmap alignment**

This extends the experimental chat connector entry points under Agent
Chat and Connected Apps. The project owner requested this feature.

Refs #11418 for the original provider catalog proposal. This
implementation uses the newer chat connection infrastructure rather than
an MCP catalog-only integration. Refs #15721 for remote OpenCode
startup.

## What Changed

- Add the Speko Apps catalog entry, provider artwork, and browser SDK.
- Reuse existing agents in setup. Keep the optional Company Phone Agent
in the catalog, with connection instructions available to any assigned
agent.
- Add task voice controls, transcripts, phone callbacks, inbound-line
settings, and call Activity entries.
- Surface retryable connection and incoming-task inventory failures,
preserve active call controls and cached tasks, restore keyboard focus
after recovery, and bound notification polling so a stuck request cannot
hide later results.
- Put Save & exit and Connect Speko in the credential step’s shared
footer.
- Add deterministic production-component Storybook fixtures for setup,
lifecycle, failures, permissions, and phone surfaces.
- Include voice components in the design guide. Stories never request a
real microphone or place calls.

## Verification

- All 27 focused UI/controller regressions, UI typecheck, token gates,
and 28 map/catalog structural tests pass. The shared connection cache
test covers pause/resume invalidation during the 30-second fresh period.
- Storybook builds 157 Speko production-component stories. The complete
prior 156-story set passed all 624 checks on UI source 249cbc616a. After
the final cache repair, all 152 affected page/launcher checks passed
across both themes, 390/1200 widths and reduced motion. Real microphone
access and external provider requests are blocked.
- Source annotations follow the current generated-catalog contract. No
generated catalog or lockfile is included. Three inherited
GitHub-heading tests passed after master reconciliation.
- Latest-head CI and Greptile review must pass before merge. Local
fixture success does not establish provider or spoken-playback
acceptance.
- The separate integrated browser run passed two journeys and failed
browser answer retrieval. Real browser/telephone playback and timing
acceptance remain outstanding. See the dated qualification report and
E01–E16 runbook. Speko stays experimental.

## Risks

- Voice sessions add migrations and provider-dependent lifecycle
behavior. Migrations preserve existing chat and AI transports.
- Callback origins must be reachable HTTPS. Provider keys and signing
secrets remain vaulted.
- Unverified callers receive task-scoped low-trust authority. Caller ID
does not grant private task access.
- Outbound telephony incurs provider charges. Provider acceptance is not
proof that the caller heard an answer.
- Keep the feature behind the existing experimental chat setting until
the real browser, inbound, outbound, and delayed-result acceptance
matrix passes.

## Model Used

OpenAI Codex, GPT-6, with tool use and code execution. The exact serving
model ID and context window are not exposed in this session.

## 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
- [ ] 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
- [x] All Paperclip CI gates are green
- [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

---------

Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-10-13 16:26:38 -05:00

102 lines
5.8 KiB
TypeScript

import { afterEach, expect, it, vi } from "vitest";
import { voiceSessionsApi } from "./voiceSessions";
import { createVoiceCallAttempt, voiceCallJournal } from "../lib/voice-call-attempt";
import { createVoiceSessionController, VoiceSessionStartError } from "../lib/voice-session-controller";
afterEach(() => { vi.restoreAllMocks(); vi.unstubAllGlobals(); });
it("aborts a stuck notification fetch so the next poll can retrieve the saved answer", async () => {
const deadline = new AbortController();
const timeout = vi.spyOn(AbortSignal, "timeout").mockReturnValue(deadline.signal);
const fetch = vi.fn().mockImplementationOnce((_url: string, init: RequestInit) => new Promise((_resolve, reject) => {
init.signal!.addEventListener("abort", () => reject(new DOMException("Aborted", "AbortError")), { once: true });
})).mockResolvedValueOnce(new Response(JSON.stringify({sessionId: "session", generation: 1, publicationId: "saved-answer", attempt: 0}), {headers: {"content-type": "application/json"}}));
vi.stubGlobal("fetch", fetch);
const stalled = voiceSessionsApi.notification("company", "session");
const rejected = expect(stalled).rejects.toMatchObject({name: "AbortError"});
expect(timeout).toHaveBeenCalledWith(10_000); deadline.abort(); await rejected;
timeout.mockReturnValue(new AbortController().signal);
await expect(voiceSessionsApi.notification("company", "session")).resolves.toMatchObject({publicationId: "saved-answer"});
expect(fetch).toHaveBeenCalledTimes(2);
});
function recoveryJournal() {
const rows = new Map<string, string>();
return voiceCallJournal({
getItem: key => rows.get(key) ?? null,
setItem: (key, value) => { rows.set(key, value); },
removeItem: key => { rows.delete(key); },
}, "company", "operator");
}
function stalledFetch(_url: string, init: RequestInit) {
return new Promise<Response>((_resolve, reject) => {
init.signal!.addEventListener("abort", () => reject(new DOMException("Timed out", "TimeoutError")), { once: true });
});
}
function jsonResponse(value: unknown) {
return new Response(JSON.stringify(value), { headers: { "content-type": "application/json" } });
}
it("bounds lost start responses, preserves the request across reload, and closes the recovered call", async () => {
const deadline = new AbortController();
const timeout = vi.spyOn(AbortSignal, "timeout").mockReturnValue(deadline.signal);
const fetch = vi.fn().mockImplementationOnce(stalledFetch)
.mockResolvedValueOnce(jsonResponse({ session: { id: "saved-session", state: "active" } }))
.mockResolvedValueOnce(jsonResponse({ id: "saved-session", state: "ended" }));
vi.stubGlobal("fetch", fetch);
const journal = recoveryJournal();
const attempt = createVoiceCallAttempt("company", voiceSessionsApi, journal);
const controller = createVoiceSessionController({
async mint() { return (await attempt.start({ endpointId: "endpoint", issueId: "task" })).media; },
async end(id) { await attempt.end(id); },
connect: vi.fn(),
});
const starting = controller.start();
const stopping = controller.stop();
expect(controller.getSnapshot().state).toBe("ending");
expect(timeout).toHaveBeenCalledWith(10_000);
deadline.abort();
await Promise.all([starting, stopping]);
expect(controller.getSnapshot()).toMatchObject({ state: "ended", busy: false });
const original = journal.read();
expect(original?.request).toMatchObject({ endpointId: "endpoint", issueId: "task" });
timeout.mockReturnValue(new AbortController().signal);
const reloaded = createVoiceCallAttempt("company", voiceSessionsApi, journal);
await expect(reloaded.start({ endpointId: "other-endpoint" })).rejects.toBeInstanceOf(VoiceSessionStartError);
expect(fetch.mock.calls[1][1].body).toBe(fetch.mock.calls[0][1].body);
expect(journal.read()?.sessionId).toBe("saved-session");
await reloaded.end("saved-session");
expect(journal.read()).toBeUndefined();
});
it("releases stalled hangup into cleanup retry and preserves the exact call identity", async () => {
const deadline = new AbortController();
const timeout = vi.spyOn(AbortSignal, "timeout").mockReturnValue(new AbortController().signal);
const media = { sessionId: "saved-session", generation: 1, transportToken: "short-lived", transportUrl: "wss://test.invalid" };
const fetch = vi.fn().mockResolvedValueOnce(jsonResponse({ session: { id: "saved-session", state: "active" }, media }))
.mockImplementationOnce(stalledFetch)
.mockResolvedValueOnce(jsonResponse({ id: "saved-session", state: "ended" }));
vi.stubGlobal("fetch", fetch);
const journal = recoveryJournal();
const attempt = createVoiceCallAttempt("company", voiceSessionsApi, journal);
const localMedia = { endSession: vi.fn(async () => {}), setMicMuted: vi.fn(), sendChatMessage: vi.fn(), startAudioPlayback: vi.fn() };
const controller = createVoiceSessionController({
async mint() { return (await attempt.start({ endpointId: "endpoint", issueId: "task" })).media; },
async end(id) { await attempt.end(id); },
async connect() { return localMedia; },
});
await controller.start();
timeout.mockReturnValue(deadline.signal);
const stopping = controller.stop();
// Cleanup launches remote and local shutdown together.
await Promise.resolve(); await Promise.resolve();
expect(timeout).toHaveBeenLastCalledWith(10_000);
deadline.abort(); await stopping;
expect(controller.getSnapshot()).toMatchObject({ state: "cleanup_failed", busy: false });
expect(journal.read()?.sessionId).toBe("saved-session");
await controller.start(); expect(fetch).toHaveBeenCalledTimes(2);
timeout.mockReturnValue(new AbortController().signal);
await controller.stop();
expect(controller.getSnapshot()).toMatchObject({ state: "ended", busy: false });
expect(fetch.mock.calls[2][0]).toBe(fetch.mock.calls[1][0]);
expect(localMedia.endSession).toHaveBeenCalledOnce();
expect(journal.read()).toBeUndefined();
});