mirror of
https://github.com/vercel/chat.git
synced 2026-09-14 18:32:29 +08:00
ac8a20779c
## Summary
Introduces a dedicated `chat/ai` subpath as the home for every Vercel AI
SDK helper that ships with Chat SDK. Importing from this subpath keeps
the optional `ai` and `zod` peer dependencies out of bundles that don't
use them.
### What's new
- **`createChatTools`** — exposes Chat SDK operations as ready-to-use AI
SDK tools so an agent can read, post, react, edit, delete, and manage
thread subscriptions across every adapter the supplied `Chat` instance
has registered.
- Write operations require user approval by default (`requireApproval:
true`); toggle globally or per-tool.
- Three presets — `reader`, `messenger`, `moderator` — scope the
toolset.
- Individual tools can also be cherry-picked (`import { postMessage,
addReaction } from "chat/ai"`).
- **`toAiMessages`** (and the `Ai*` / `ToAiMessagesOptions` types) now
live alongside the tools at `chat/ai`. The previous `chat` re-exports
continue to work, but are flagged `@deprecated` with an editor hint
pointing to the new home — migration is a one-line import change.
- **Docs** — new `/docs/ai` section between Usage and Adapters in the
sidebar:
- `/docs/ai` — Overview
- `/docs/ai/ai-sdk-tools` — `createChatTools` guide
- `/docs/ai/to-ai-messages` — `toAiMessages` reference
- `/docs/ai/types` — Reference for every type exported from `chat/ai`
- **Example app** — `examples/nextjs-chat` now demos the new surface via
a "Run Agent Demo" button on the welcome card and a free-form `/agent
<prompt>` slash command (streaming, with a placeholder so users get
immediate feedback in channel contexts where Slack's typing-status API
is a no-op).
### Future plans
`createChatTools` currently exposes the cross-adapter Chat SDK surface
only. A natural follow-up is to also support **platform-specific tools**
— e.g. expose Slack-only `pin`/`unpin`, Discord-only thread archiving,
GitHub-only issue commenting, etc., so users can further extend what
their agent can do without dropping back to raw adapter calls. The shape
would likely be additional opt-in factories under `chat/ai` (or
per-adapter subpaths like `@chat-adapter/slack/ai`) that return tools
layered on top of the platform-specific adapter clients, while keeping
the cross-platform `createChatTools` API as the lowest common
denominator.
### Coverage
- `createChatTools` orchestrator: 100% statements / 94.7% branches.
- Every tool factory's `execute()` is exercised end-to-end (29 tests in
`index.test.ts`).
- `toAiMessages` keeps its existing 35-test suite covering role mapping,
attachment handling, links, transforms, and unsupported-attachment
fallbacks.
- Tools folder overall: 99.0% statements / 86.1% branches / 97.4%
functions / 98.9% lines.
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Co-authored-by: dancer <josh@afterima.ge>
Integration Tests
Integration tests for the Chat SDK that verify real-world webhook payloads are handled correctly.
Test Categories
- Unit tests (
slack.test.ts,teams.test.ts,gchat.test.ts) - Test adapter functionality with mock payloads - Replay tests (
replay*.test.ts) - Replay actual production webhook recordings - Emulator tests (
src/emulator/<adapter>/*.test.ts) - Drive the SDK against an in-process Emulate.dev server, one per supported adapter (@emulators/slack,@emulators/github). Assertions read the emulator's stateful store (messages, comments, reactions, installations) instead of mock call records. Each adapter is wired in via itsapiUrlconfig. Inbound deliveries: the Slack flow re-signsevent_callbackpayloads withx-slack-signaturevia a small in-test forwarder before handing them tochat.webhooks.slack(...), while the GitHub flow is a near-passthrough because the emulator'sWebhookDispatcheralready signs withX-Hub-Signature-256exactly as the adapter expects. Helpers live insrc/emulator/slack/utils.tsandsrc/emulator/github/utils.ts.
Replay Tests
Replay tests use recorded webhook payloads from production to verify the SDK handles real interactions correctly.
See fixtures/replay/README.md for:
- How to record new fixtures
- Fixture format documentation
- SHA-based recording workflow
- Platform-specific webhook formats
Running Tests
# Run all integration tests
pnpm --filter @chat-adapter/integration-tests test
# Run with watch mode
pnpm --filter @chat-adapter/integration-tests test:watch