## Summary
Upgrades the chat-sdk.dev docs app from `@vercel/geistdocs` **1.20.4 →
1.22.0** (published 2026-08-21, Apache-2.0) and `next` **16.2.11 →
16.3.1**, following the bundled 1.22.0 template as the source of truth.
The target release includes all of the behavior-changing PRs for this
round:
- vercel/geistdocs#245 — require Next.js 16.3, scaffold 16.3.1, drop the
dev filesystem-cache flag (shipped in 1.21.1)
- vercel/geistdocs#246 — Cache Components across Geistdocs
- vercel/geistdocs#249 — Partial Prefetching + instant docs navigation
- vercel/geistdocs#250 — stable Next 16.3 APIs, retryable page/Ask AI
error boundaries, full prefetch of package links, no generic page shell
- vercel/geistdocs#251 — tree sidebar preserves scroll position on
folder toggles
## Adapter and configuration changes
- `next.config.ts`: `cacheComponents: true`, `partialPrefetching: true`;
removed `experimental.turbopackFileSystemCacheForDev` (default in 16.3).
Redirects, `/sitemap.xml` rewrite, and image config unchanged.
- New `lib/geistdocs/root-params.ts`; all layouts read `[lang]` via
`next/root-params` instead of `params`. Route handlers keep
route-context `params`.
- Root layout gains `generateStaticParams` returning every configured
language (`en`) — home (`/`), `/adapters`, and `/resources` now
prerender statically (previously dynamic).
- Route adapters no longer re-export `revalidate`/`dynamic` from package
factories (`agents.md`, `sitemap.md`, `llms.mdx`), and custom routes
drop their own `revalidate` exports (`llms.txt`, `llms-full.txt`,
`adapters.mdx`, `rss.xml`, `resources`).
- `llms.mdx` adopts the template form: `sources: [geistdocsSource]` +
`notFound: {}`, enabling smart agent-readable 404/410 responses with
real HTTP statuses.
- `rss.xml` migrated to the template's `"use cache"` +
`cacheLife("max")` form with `getPublicPath` base-path handling.
- `resources` page: `revalidate = 86400` → `"use cache"` +
`cacheLife("days")` (same 1-day lifetime).
- App-owned data fetching moved off `next: { revalidate }` (unsupported
under Cache Components): GitHub README fetches and homepage OSS stats
now use `"use cache"` + `cacheLife("hours")`.
- Homepage Shiki highlighting (`Demo`, `CodePreview`, `highlightCode`)
runs inside `"use cache"` scopes — Shiki reads `Date.now()` internally,
which otherwise fails prerendering.
- App-owned links to statically generated docs pages get
`prefetch={true}` (platform grid, feature matrix, adapter slug list,
"Visit Documentation") per the template's agent guidance; package-owned
sidebar/prev-next links already prefetch fully in 1.22.0.
- `Analytics`/`SpeedInsights` moved into
`components/geistdocs/provider.tsx` per the template.
- CSS: `styles/geistdocs.css` now imports `@vercel/geistdocs/theme.css`
(self-sourcing package dist/streamdown) instead of layering on
`styles.css` from `global.css`; updated the mobile breadcrumb selector
for the new package DOM; kept the site-specific shadcn tokens, dark
background-scale override, prose, TOC, and streamdown fixes.
- Added `apps/docs/AGENTS.md` capturing the packaged-architecture
conventions (cache-components rules, root-params, markdown contract,
proxy mappings).
## PR #251 (tree sidebar scroll) verification
The fix is package-internal (`manualToggleRef` in `SidebarTree`); no
consumer change is needed. This site's sidebars render no collapsible
folder rows (content uses spread folders, `...api` etc.), so I verified
the shipped behavior against the bundled 1.22.0 template with
`sidebarMode="tree"` enabled locally: expanding/collapsing a folder
preserves the exact sidebar scroll position (772 → 772), and a route
change into a collapsed folder still scrolls the active item into view.
8/8 checks pass.
## Static-generation coverage
Production build: 290/290 static pages generated. Every intended
parameter tuple is prerendered with complete content (verified H1/body
in emitted HTML):
- `/en` home, `/en/adapters` listing, `/en/resources` — now fully static
(were `ƒ` on main)
- `/en/docs/*` — 45 pages, complete static HTML + one generic
`[[...slug]]` fallback entry (allowed)
- `/en/adapters/{official,community,vendor-official}/*` — 44 detail
pages + `/en/adapters.mdx/*` markdown for all 44
- `/en/sitemap.md` — SSG
Intentional contract differences (match the 1.22.0 template's own build
output):
- OG image routes (`/og/[...slug]`, adapter `*/og`) render on demand
under Cache Components instead of build-time SSG; Next 16.3 caches the
rendered image per route. URLs and content types verified unchanged.
- `llms.txt`, `llms-full.txt`, `llms.mdx`, `rss.xml`, `agents.md` remain
on-demand route handlers (same as main); `agents.md` reads the request
origin by package design.
- Unknown HTML routes: browsers receive the docs shell with 200 before
not-found UI resolves; crawlers get a real 404. Machine-readable unknown
routes return the new smart 404 body with real 404 status and
`X-Robots-Tag: noindex`.
## Lockfile
`pnpm-lock.yaml` delta: the `docs` importer's `@vercel/geistdocs`
(1.20.4 → 1.22.0) and `next` (16.2.11 → 16.3.1) bumps, their
peer-context re-resolutions, and one mechanical re-keying of `@swc/core`
peer contexts to include `@swc/helpers` across existing entries (no
version changes outside the docs app). Verified with `pnpm install
--frozen-lockfile`.
## Test results
- `pnpm install --frozen-lockfile` ✓
- `pnpm check` ✓ (1 pre-existing warning in untouched
`lib/read-more.ts`)
- `pnpm typecheck` — 43/43 ✓
- `pnpm knip` ✓, `pnpm konsistent` ✓
- `pnpm test` — 47/47 turbo tasks ✓
- Clean production build (removed `.next`/`.source`) ✓ 290/290
- Production-server (`next start`) contract checks: HTML docs,
`.md`/`.mdx`, `Accept: text/markdown`, agent-UA negotiation, `llms.txt`,
`llms-full.txt`, `sitemap.md`, `agents.md`, `/.well-known/mcp.json`
(intentional 404), `rss.xml`, `robots.txt`, `/sitemap.xml` rewrite, OG
images, `AGENTS.md`, all redirects (308s), `/api/search` JSON and not
rewritten as markdown, `/docs.md` section root ✓
- Browser checks (Playwright, Chrome, `next start`) — 16/16: instant
sidebar + prev/next client navigation with complete content and no
loading shell, single visible H1 (Activity-preserved routes stay
hidden), Copy Page, page actions menu, search → result navigation, Ask
AI panel with suggestions, theme switch (dark applies the site's
background scale), mobile navbar menu and docs sheet navigation, unknown
page shows not-found UI, zero console/page errors and failed requests
- Ask AI scoped failure: with no local AI Gateway credentials the chat
surfaces the error, the surrounding page stays intact, and resubmission
retries cleanly
- Visual parity screenshots vs production (home/docs/adapters, light +
dark) match
## Preview
- Preview:
https://chat-git-richardhaines-geistdocs-122-upgrade.vercel.sh —
deployment **Ready**, build passed on Vercel.
- The preview sits behind Vercel SSO (Fork Protection), so automated
route checks aren't possible without a bypass token; please spot-check
through SSO: `/docs/getting-started`, `/docs/getting-started.md`,
`/adapters`, `/llms.txt`, and a client navigation between docs pages.
Signed-off-by: molebox <rich@vercel.com>
Many docs pages are orphaned: nothing links to them apart from the
sidebar, so readers and crawlers rarely find them. This PR gives every
docs page a Read more section with four cards at the bottom of the
article, above the prev/next footer.
Cards are picked deterministically in lib/read-more.ts: the page's
related frontmatter first, then prerequisites, then siblings from the
same sidebar section, then the rest of the page tree, so every page
always fills all four slots. Card titles and descriptions come from the
target page's own frontmatter, nothing is duplicated. The section is
injected through the MDX wrapper slot in the docs route, so it applies
to all pages without touching content.
To make the links topical rather than positional, 26 pages get related
frontmatter additions. The 20 pages that no other page referenced (all
ten api/ pages among them) now each have at least one inbound link,
generally pairing guides with their API reference and back. The bundled
copy of create-chat-sdk.mdx is synced to keep the byte-match test green.
Official adapter pages get the same treatment with a More adapters
section: same-type adapters first (platform or state, using the catalog
order), topped up from the other official group. Vendor-official and
community adapters are never shown, and their pages don't render the
section. It reuses AdapterCard, so logos and package names match the
listing page.
Two small SEO fixes ride along. JSON-LD was allowlisted to three docs
pages; the allowlist is gone, so all 45 now emit HowTo or TechArticle
plus a BreadcrumbList. Docs and adapter detail pages also emit canonical
URLs now, resolved against the existing metadataBase.
Verified against the production build: all 45 docs pages and all 19
official adapter pages render exactly four cards, no page is left
unreferenced, canonicals and JSON-LD are present everywhere, and pnpm
validate passes. Docs-only, so no changeset.
---------
Signed-off-by: Ben Sabic <bensabic@users.noreply.github.com>
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
## Problem
On the GetStarted cards in the home page, the `<pre>` inside
`CodePreview` had `bg-background` hardcoded. When hovering a card
(`hover:bg-muted/40`), the code block retained its own opaque
background, so the hover tint only showed through around the text — the
dark/muted fill appeared clipped to the text container rather than
filling the whole rotated card.
## Fix
Removed `bg-background` from the `<pre>` element so it inherits the
parent card's background. The default state is unchanged since the card
itself already has `bg-background`.
---------
Co-authored-by: v0 <it+v0agent@vercel.com>
Co-authored-by: Matias Gonzalez <29680544+matiasngf@users.noreply.github.com>
Adds `@chat-adapter/notion`, an official adapter that lets a Chat SDK
bot take part in **Notion comment discussions** (page-level and
block/discussion threads) with the same handler code used for Slack,
Linear, GitHub, etc. Inbound events arrive via Notion webhooks
(`comment.created`) with HMAC signature verification; outbound actions
use the Comments REST API. Because Notion lets a connection edit its own
comments, the adapter supports **Post+Edit streaming**.
### Highlights
- **Webhooks** — `comment.created` verified with `X-Notion-Signature`
HMAC over the raw body (timing-safe), plus the one-time
`verification_token` handshake. Returns a fast 200 with idempotent,
state-backed dedupe.
- **Post+Edit streaming** — posts the first chunk, then `PATCH`es the
comment as tokens arrive, throttled to Notion's ~3 req/s limit (global
token bucket, `Retry-After` aware). Long bodies are split into
sequential comments to stay under the 2000-char rich-text cap.
- **Mentions** — three modes: `mention` (default; plain-text `@userName`
/ `@botUserId`), `all-comments`, and `keyword`.
- **`message.subject`** — resolves the parent page via the Pages API
(title, url, archived status, author).
- **File uploads** — up to 3 native attachments via the File Uploads API
(binary `single_part`; public URLs via `external_url` with bounded
polling); overflow and failures fall back to markdown links.
- **History** — `fetchMessages` over list-comments (open comments only),
direction-aware.
- Cards render as markdown fallback; reactions / typing / DMs are typed
no-ops or errors. Registered in the `chat/adapters` catalog and the
`create-chat-sdk` scaffold; pinned to `Notion-Version: 2026-03-11`.
### Usage
```ts
// lib/bot.ts
import { Chat } from "chat";
import { createNotionAdapter } from "@chat-adapter/notion";
import { createRedisState } from "@chat-adapter/state-redis";
export const bot = new Chat({
userName: "notion-bot",
adapters: { notion: createNotionAdapter() }, // reads NOTION_TOKEN + NOTION_VERIFICATION_TOKEN
state: createRedisState(),
});
bot.onNewMention(async (thread, message) => {
const subject = await message.subject; // parent page metadata (title, url, …)
await thread.post(`Thanks for the mention on **${subject?.title ?? "this page"}**!`);
});
```
```ts
// app/api/webhooks/notion/route.ts
import { bot } from "@/lib/bot";
export const POST = (request: Request): Promise<Response> => bot.webhooks.notion(request);
```
### Configuration
Auto-detects `NOTION_TOKEN` and `NOTION_VERIFICATION_TOKEN`, plus
optional `NOTION_BOT_USERNAME`, `NOTION_MENTION_MODE`,
`NOTION_KEYWORDS`, and `NOTION_VERSION`; everything is overridable via
`createNotionAdapter({ … })`. The docs page covers the full connection +
webhook setup (capabilities, content access, and the webhook-URL-lock
warning).
Changeset bumps `@chat-adapter/notion`, `chat`, and `create-chat-sdk`
(minor). Layered as four commits: `feat` (adapter +
catalog/scaffold/emoji), `docs`, `test`, `chore(example)`.
---------
Signed-off-by: Ben Sabic <bensabic@users.noreply.github.com>
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Co-authored-by: dancer <josh@afterima.ge>
Adds a vendor-official contributing guide covering qualifications,
listing terms, and the PR checklist for platform vendors. Contributing
and adapter overview pages point to that guide for listing details
instead of repeating them.
Moves Slack and Teams low-level API docs onto their adapter pages, with
permanent redirects from `/docs/slack-primitives` and
`/docs/teams-primitives`.
Trims stale hand-maintained comparison tables from the docs intro and
platform adapters overview. Those pages now link to `/adapters` and the
generated official feature matrix. Adds contributing CTAs for building
an adapter and listing a vendor-official one.
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
- Add a dedicated XChat speech-bubble logo for the docs hero and
`/adapters` card
- Point XChat docs and `adapters.json` at the new `xchat` icon instead
of reusing `x`
- Update the XChat OG image
- Add reciprocal “X Adapter vs XChat Adapter” / “XChat Adapter vs X
Adapter” sections on both docs pages
- Rename remaining “X Chat” references to “XChat” in the adapter package
README and comments
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Third repo in the design sync, after `vercel/geistdocs#216`/`#218` and
`vercel/flags#457`. Upgrades the docs site to `@vercel/geistdocs@1.19.2`
and brings the homepage onto the Geist design system.
`apps/docs` is `private: true`, so no changeset.
## Dependency
`1.19.2` peers on `next: ^16.2.11` and the app pinned `16.2.6`, so
**next moves to `16.2.12`** alongside it — without that pnpm reports an
unmet peer. Installed via `pnpm add --save-exact` per AGENTS.md. All 20
geistdocs subpaths this app imports still exist in 1.19.2; no API
breakage.
The footer needed no work: 1.16 already shipped the prop-less
Vercel-directory `<Footer />`.
## Layout — `home-grid.css` is gone
Deleted `app/styles/home-grid.css` (368 lines) and its `global.css`
import, and rebuilt each section on `grid-cols-12` / `col-span-*`:
| section | before (CSS) | after |
|---|---|---|
| OSS stats | 2×2 → 4×1 @768 | `col-span-6 min-[768px]:col-span-3` |
| Features | 2-up + full-width 3rd → 3×1 @961 | `col-span-12
sm:col-span-6 lg:col-span-4` |
| Code | stacked → sidebar 1/3 + code 2/3 @961 | `lg:col-span-8` /
`lg:col-span-4`, pinned with `col-start` + `row-start` |
| Integrations | 1×5 → tall left + 2×2 @961 | `lg:col-span-4
lg:row-span-2` + four `lg:col-span-4` |
The code section needs explicit `col-start`/`row-start` because the
sidebar follows the code in the DOM but sits left of it from `lg`.
Other layout changes:
- **Single gutter at the page root** (`mx-auto w-full max-w-[1448px]
px-4 sm:px-6`); removed the per-section horizontal padding that
duplicated it, so every section's content lands on the navbar/footer
content edge.
- **Content widened 1114px → 1400px**, the navbar's content span (1448 −
2×24).
- **Bottom gap above the footer trimmed ~320px → ~176px** — layout
`pb-32` → `pb-16` and page `pb-24 sm:pb-36` → `pb-12 sm:pb-16`. Three
paddings were stacking.
The 768px stats breakpoint is preserved with `min-[768px]:` — there's no
Tailwind equivalent here (`md`=601, `lg`=961) and four KPI columns at
601px would be ~140px each.
## Design — ported from vercel.com/ai-sdk
Read off the flagged source in `front/apps/vercel-marketing/.../ai-sdk`,
not the live site.
- **Code showcase tabs** → the `SlidingTabs` primitive: pill labels with
an animated indicator, full keyboard nav (arrows/Home/End, roving
tabindex), and an invisible-bold label so the tab doesn't shift width
when it bolds. Four tabs per group with dot pagination for the rest,
tabs above the code block. Copied into `components/ui/sliding-tabs.tsx`
with `cn` rewired and the `no-scrollbar` utility inlined (geistdocs
doesn't define it).
- **"Scale with confidence"** → heading and paragraph on one
bottom-aligned row (cols 1–4 / 8–12), then four bordered cards
`col-span-12 md:col-span-6 lg:col-span-3`. Type mapped from their
primitives: `SectionHeading size="48"` → `text-heading-40
lg:text-heading-48`, `SectionParagraph size="18"` → `text-copy-16
lg:text-copy-18`.
- **Feature row** → icon + muted eyebrow over a prominent statement.
Note this **inverts the previous emphasis**: the heading is now the
small muted label and the description the larger line, matching the
reference. Icons come from geistdocs' own set so they match Geist's line
weight: `IconLinked`, `IconWorkflow`, `IconAcronymTs`.
- **Get-started install snippet** → the shared `CommandPrompt`, with its
buttons on one row from `lg`.
- Remaining headings converted to `text-heading-*`.
- Navbar logo drops `height={22}` to take `LogoChatSdk`'s new 18px
default (renders 106.9×22 → 87.4×18).
## Two fixes worth calling out
**`lib/utils.ts` — `cn` was silently dropping typography.** Geist's
`text-copy-*`/`text-heading-*` share the `text-` prefix with colour
utilities, so stock `tailwind-merge` classifies them as colours and
drops the size whenever both appear in one `cn()` call. geistdocs ships
a `cn` that registers them as `font-size` for exactly this reason but
doesn't export it, so the config is replicated here. This was a latent
bug across the app, not just the new code.
**`Analytics`/`SpeedInsights` moved out of the `"use client"` provider**
into the server layout. Both emit `<script>`, and scripts rendered
inside a client tree never execute — so analytics wasn't firing on
client navigations. React 19.2.7 (pulled in by this bump) now warns
about it; the bug predates it.
## Verification
- `pnpm --filter docs build` passes (270 pages), `tsc --noEmit` clean,
`biome check` clean.
- Rendered output spot-checked for the tab strip, dot pagination, card
classes, and feature icons.
**`pnpm validate` could not be run** — it needs Node ≥20.19 and this
machine is on v20.11.1 (`pnpm check` dies on `styleText` from
`node:util`). Biome, tsc and build were run directly instead, but the
knip and test legs are unrun and should be confirmed in CI.
Signed-off-by: christopherkindl <53372002+christopherkindl@users.noreply.github.com>
Upgrades the docs to `@vercel/geistdocs@1.16.0`.
- Bump `@vercel/geistdocs` 1.15.5 → 1.16.0
- Drop the removed `config` prop from `<Footer />` (1.16.0 replaces the
footer with the Vercel product directory and no longer accepts props)
- Load Geist Sans from the `geist` npm package so the new `ss11`
stylistic set (alternate "I") renders — Google Fonts strips it
## Summary
Site search on chat-sdk.dev never returned adapter pages — e.g.
searching `imessage` found nothing even though several
`content/adapters/vendor-official/` pages (Sendblue, Linq, Photon, Dial)
mention iMessage prominently.
The cause: the search route only passed the `content/docs` collection
(`geistdocsSource`) to `createSearchRoute`, so the separate
`content/adapters` collection (`adaptersSource`) was never indexed by
Orama.
This adds `adaptersSource` to the route's sources. The adapters loader
already shares the same `i18n` config, and `createSearchRoute` supports
plain fumadocs loaders, so no other changes are needed.
## Testing
Ran the docs dev server and queried `/api/search` directly:
- `?query=imessage` → 46 results across 6 adapter pages (sendblue, linq,
photon, dial, agentphone, blooio), with content-level matches and
highlights
- `?query=sendblue` → returns the Sendblue page
- `?query=webhook` → core `/docs` pages still returned alongside
adapters (existing search unaffected)
- No duplicate result IDs (no per-locale double indexing)
`pnpm check` and `tsc --noEmit` in `apps/docs` are clean. Docs-only
change — no changeset.
Signed-off-by: Ben Sabic <bensabic@users.noreply.github.com>
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
## Summary
Migrates `apps/docs` from locally-copied geistdocs runtime code to the
published
[`@vercel/geistdocs`](https://www.npmjs.com/package/@vercel/geistdocs)
package (1.8.2), following the official [migration
guide](https://preview.geistdocs.com/docs/migration). Net **−8,400
lines**.
### Package-backed now
- Docs page + layouts: `createDocsPage`, `GeistdocsDocsLayout`,
`GeistdocsHomeLayout` (JSON-LD + sr-only markdown hints preserved via
`renderTop`)
- Navbar (OSS product switcher via `navbarOssProducts`), footer,
provider, search dialog, page actions (edit source, feedback, copy page,
Ask AI, open-in-chat, scroll top)
- `/api/search` → `createSearchRoute`, `/api/chat` → `createChatRoute`
(AI SDK v6; AI Gateway default, optional `GEISTDOCS_CHAT_PROXY_URL`)
- `llms.mdx` → `createDocsMarkdownRoute`, `sitemap.md` →
`createSitemapMarkdownRoute` (now includes an **Adapters** section)
- **New**: `/agents.md` via `createAgentsRoute`, backed by a new `agent`
readiness config
- `proxy.ts` → `createProxy` with explicit `markdownRoutes` for `/docs`
→ `llms.mdx` and `/adapters` → `adapters.mdx` (adds AI-agent UA
rewrites)
- CSS: `@vercel/geistdocs/styles.css` + slim local overrides (shadcn
tokens for remaining `components/ui`, body tint, prose inline code,
`#nd-*` tweaks); code blocks now use the geist Shiki theme
- Icons/logos from `@vercel/geistdocs/assets/*`; feedback via the
package action (same geistdocs.com endpoint + `siteId`)
### Kept local by design
- Curated `/llms.txt` index + `/llms-full.txt` corpus — the published
`AGENTS.md`/SKILL.md artifacts and integration tests reference this
exact contract
- The adapters section (README fetching, OG images, JSON-LD, feature
matrices, `adapters.mdx` markdown route) — now rendered inside the
package docs layout
- RSS and OG image routes (app-owned per the migration guide)
- Skipped `/.well-known/mcp.json`: no MCP servers configured, and the
proxy matcher must keep excluding `.well-known` for the served
agent-skills files
### Cleanup
- Deleted local copies: `components/geistdocs/*` chrome,
`components/ai-elements/*`, chat hooks/persistence, feedback server
actions, unused shadcn primitives, geistcn logo/icon fallbacks covered
by package assets
- Removed 13 now-unused deps (`ai@5`, `@ai-sdk/react@2`, `dexie`,
`jotai`, `cmdk`, `vaul`, `mermaid`, `nanoid`, `react-player`,
`use-stick-to-bottom`, `@orama/tokenizers`, `dexie-react-hooks`,
`next-themes`)
- Updated `docs-llms.test.ts` proxy assertions to the `createProxy`
markdown-route shape
### Behavior changes to be aware of
- Code blocks use the geist Shiki theme instead of GitHub light/dark
- Ask AI history is no longer persisted in IndexedDB (package owns the
panel)
- Adapters sidebar uses the standard geistdocs tree rendering instead of
the bespoke grouped sidebar
- Per-page markdown output appends the standard geistdocs footer links
(`/sitemap.md`, `/llms.txt`, `/agents.md`)
## Test plan
- `pnpm validate` green (knip + check + typecheck + test + build)
- Smoke-tested against `next build && next start`: `/`, `/docs`,
`/adapters`, `/agents.md`, `/llms.txt`, `/llms-full.txt`, `/sitemap.md`,
page-level `.md` URLs for both docs and adapters, `Accept:
text/markdown` negotiation, search API, JSON-LD, sr-only markdown hints,
edit-source URLs (`apps/docs/content/docs/{path}`), OSS navbar, page
actions
- Verified compiled CSS chunks contain the home grid, Shiki palette, and
geist utilities (note: stale turbopack dev caches from before this
change can serve incomplete CSS — `rm -rf apps/docs/.next` fixes it)
## Checklist
- [x] All commits are signed and verified
- [x] All commits are signed off for the DCO (`git commit -s`)
- [x] `pnpm validate` passes
- [x] Changeset added (or N/A — docs app + tests only, no package
behavior change)
- [x] Documentation updated (or N/A)
---------
Signed-off-by: molebox <rich@vercel.com>
## summary
- swap chat model strings to `xai/grok-4.5` across the docs site, the
shipped guides, and the example bots, so the docs lead with the latest
model
- apps/docs: ai overview, ai-sdk-tools, and streaming pages, the two
landing-page code samples, and the live chat demo route
(`app/api/chat/route.ts`, previously `openai/gpt-4.1-mini`)
- packages/chat/resources/guides: the seven guides that use a chat model
(slack connect, slack + ai sdk, liveblocks, vercel blob, github code
review, daily digest, ai gateway)
- examples: nextjs-chat and nuxt-chat bots
- deliberately left non-chat model strings as they were, since grok-4.5
cannot fill those roles: the `openai/text-embedding-3-small` embedding
model and the `openai/gpt-4o-mini` reranker
- in the ai gateway guide, grok-4.5 is now the primary model but the
fallback list stays cross-provider (`anthropic/claude-opus-4.8`,
`google/gemini-3.1-pro-preview`) so the failover example still
demonstrates real cross-provider fallback
- also fixed a pre-existing prose/code mismatch in that guide, the
fallback prose said `claude-opus-4.7` while the code listed `4.8`, now
aligned to `4.8`
---------
Signed-off-by: dancer <josh@afterima.ge>
## What
- Add the new **X** adapter logo to the homepage supported-platforms
grid, placed after Messenger.
- Cap the grid width on mobile (`max-w-64`, reset with `sm:max-w-none`)
so the logos wrap **4 / 4 / 3** instead of orphaning a single logo
(Telegram) on its own line.
## Why
We now ship an `@chat-adapter/x` adapter, so X belongs in the homepage
lineup. With 11 logos, the previous flex-wrap fit 5 per row on phones,
producing a lonely 5 / 5 / 1 break. Constraining the container forces
even, centered rows down to 320px-wide screens.
## Notes
- The `x` logo SVG and the `/adapters/official/x` page already exist
(added in a prior change from `main`); this only wires X into the
homepage grid.
- Docs-only change — no changeset required.
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
## summary
new `@chat-adapter/x` adapter for X (Twitter), built on the X API v2 and
the X Activity API. write bot logic once and reply to mentions, hold DM
conversations, post from the account, and like posts, like the other
Chat SDK adapters
what it supports:
- reply to public mentions (`post.mention.create`) and top-level posts
via `channel.post`
- send and receive direct messages (`dm.received` / `dm.sent`)
- edit and delete owned posts, delete own DM events
- likes as the only reaction (`emoji.heart` or `"like"`)
- buffered streaming: accumulates an LLM stream and posts once instead
of post+edit churn on a public timeline
- OAuth 2.0 user context with managed token refresh (rotating refresh
token persisted in the state adapter, optional AES-256-GCM encryption)
- webhook CRC and `x-twitter-webhooks-signature` verification
key design decisions:
- DMs are threaded by the other participant's user id (`x:dm:{userId}`)
because X DM webhooks carry no conversation id, only participants
- OAuth 2.0 only at runtime: DM send and read are verified to work on
OAuth 2.0 user tokens, so no OAuth 1.0a in the adapter (subscription and
webhook setup is one-time and handled in the X developer console)
- parsers were written against real captured payloads: mentions use the
v2 shape (author hydrated in `includes.users`), DMs use the legacy
Account Activity shape (`direct_message_events`,
`message_create.message_data`, a `users` map, and no conversation id)
also includes the `chat/adapters` catalog entry, docs page, CLI scaffold
spec, and `sample-messages.md` with real captured payloads
<details><summary>usage</summary>
```typescript
import { Chat } from "chat";
import { createXAdapter } from "@chat-adapter/x";
const bot = new Chat({
userName: "mybot",
adapters: { x: createXAdapter() },
});
bot.onNewMention(async (thread, message) => {
await thread.post(`hi @${message.author.userName}!`);
});
bot.onDirectMessage(async (thread) => {
await thread.post("hello from X");
});
```
</details>
## test plan
- adapter unit tests pass against the real captured payload shapes, with
regression tests for author-from-`includes` (mentions) and the legacy
`direct_message_events` shape (DMs)
- real captured `post.mention.create` and `dm.received` payloads
verified end-to-end through `handleWebhook`: signature verification,
routing, author resolution, and participant threading, plus
bad-signature rejection returns 401
- every write and read path fired live against the X API through the
adapter: top-level post, reply to a mention, like and unlike, edit,
delete, DM send, DM read, DM delete
- OAuth 2.0 managed token refresh exercised live (access and refresh
token rotation)
---------
Signed-off-by: dancer <josh@afterima.ge>
Adds a Vercel Connect guide to the docs under **Usage**
(`chat-sdk.dev/docs/vercel-connect`), covering connector setup, trigger
forwarding, the per-platform `connect*Adapter` helpers from
`@vercel/connect/chat`, custom OIDC webhook verification and its trust
boundary, and limitations. Also adds a "Vercel Connect Guide" card to
the docs homepage.
This is the base of a stack; the adapter, tests, and CLI PRs below build
on it. Docs-only, so no changeset.
## Stack
- #647 — feat(chat): Vercel Connect (this PR, base → `main`)
- #648 — docs(slack): document Vercel Connect support
- #649 — feat(linear): add Vercel Connect support
- #650 — feat(github): add Vercel Connect support
- #654 — feat(tests): add Vercel Connect webhook contract helper
- #655 — feat(create-chat-sdk): add Vercel Connect mode
All of the above are stacked on this branch (`vercel-connect/base`).
## Companion
The helpers this documents ship in the `@vercel/connect/chat` subpath:
vercel/vercel#16826.
---------
Signed-off-by: Ben Sabic <bensabic@users.noreply.github.com>
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
- Break hero title cleanly on desktop only so "for" doesn't strand on
mobile
- Keep OSS stat labels (e.g. "Weekly downloads") on one line
- Shrink feature box headings on mobile, full size from sm
- Span the third feature box full-width on mobile
- Enable horizontal scroll for long code lines in the code showcase
- Match "more adapters" link color to the "Supports" label
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
## summary
restores Open Graph and Twitter preview images across the homepage,
adapters, and resources pages
adds the canonical metadata base so social image URLs resolve against
chat-sdk.dev instead of localhost
Stacked on top of #603 (`create-chat-sdk`).
Refreshes the chat-sdk.dev homepage styling:
- Adds Geist typography utilities (`text-heading-*`, `text-copy-*`) and
applies them to the hero, section headings, and copy.
- Adds a grid-based layout (`home-grid.css`) with consistent guide lines
for the stats, supported-platforms, code, and integrations sections.
- Adds a tabbed code showcase for the Chat SDK Core section, with window
chrome, a Geist syntax theme, and
`bot.ts`/handlers/cards/streaming/tools/state/multi-platform snippets.
- Scopes inline-code styling and sets the dark-mode
`--ds-background-100`/`--ds-background-200` tokens so `background-100`
is the elevated surface.
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Adds `create-chat-sdk`, a CLI that scaffolds a Next.js Chat SDK bot
project:
```bash
npm create chat-sdk@latest my-bot
# non-interactive
npm create chat-sdk@latest -- my-bot --adapter slack redis -y
```
The user picks platform and state adapters interactively or via
`--adapter`, and the CLI generates a webhook-only project with
`src/lib/bot.ts`, `.env.example`, `next.config.ts`, `package.json`, and
a README, then optionally runs `git init` and installs dependencies.
There are no pages or client UI in the template.
Adapter choices come straight from the `chat/adapters` catalog, so the
CLI has no adapter registry of its own. When a coding agent such as
Cursor or Claude Code runs the CLI, it uses non-interactive defaults and
requires an explicit platform adapter. `--interactive` forces prompts.
## also in this pr
- `google-chat` is renamed to `gchat` everywhere, including docs pages,
the OG image, and adapter catalog. Old URLs redirect permanently,
including language-prefixed and `/og` paths
- a new docs page is available at `chat-sdk.dev/docs/create-chat-sdk`,
and the CLI is promoted on the homepage, package READMEs, and agent
skill
- `create-chat-sdk` releases independently with a minor changeset for
its initial `0.1.0` release
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Co-authored-by: dancer <josh@afterima.ge>
Adds official platform adapter branding across the docs site, npm
READMEs, and social previews.
- **Homepage**: add Twilio to the supported-platforms logo grid
- **OG images**: add custom artwork for all 11 official platform
adapters under content/adapters/official/og/; serve static-first from
the existing /adapters/official/{slug}/og route with dynamic fallback
for state adapters
- **READMEs**: add linked hero banners to every official platform
adapter package README, using the live OG URL as the single image source
- **Tests**: integration-test guardrails for OG image coverage, README
banner discoverability, and knip-clean helpers
- **Changeset**: empty changeset for CI
<img width="2400" height="1256" alt="Chat SDK - Slack"
src="https://github.com/user-attachments/assets/4d186a1c-5651-44b8-8698-091ee23b44da"
/>
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Improves Chat SDK discoverability across npm, READMEs, and the docs site
for search engines and AI coding agents.
- **npm metadata**: point every published package `homepage` at
chat-sdk.dev deep links; expand `chat` keywords/description; fix
`repository.directory` (`packages/chat-sdk` → `packages/chat`); align
state adapter keywords
- **READMEs**: add npm callouts, Documentation/Guides links, and AI
Coding Agents sections (skill install, optional Vercel Plugin,
`llms.txt` / `llms-full.txt`) across all published packages and the repo
root
- **docs JSON-LD**: `HowTo` / `TechArticle` on getting-started,
streaming, and cards; `CollectionPage` + official-only `ItemList` on
`/adapters` (with split human vs JSON-LD descriptions)
- **UTMs**: add `chat-sdk_site` / `chat-sdk_repo` tracking params to
Resources links in selected MDX pages and adapter READMEs (discord,
github, slack, liveblocks, getting-started, ai index)
- **contract tests**: integration-tests guardrails for npm metadata and
README discoverability so future package additions don't drift
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Adds Schema.org JSON-LD to every adapter detail page (official,
community, vendor-official):
- `SoftwareSourceCode` describing the package — name (package),
description (tagline), repository, `programmingLanguage: TypeScript`,
`runtimePlatform: Node.js`, and author (Vercel for official adapters,
the vendor/community author otherwise). License is only emitted for
official adapters; no fabricated ratings/offers.
- `BreadcrumbList` for the `Chat SDK › Adapters › <name>` hierarchy.
Built from existing frontmatter + `adapters.json` via a shared
`getAdapterJsonLd` helper, injected with the same `<script
type="application/ld+json">` pattern already used on the home page.
Primarily a classic-SEO/rich-result and entity-linking improvement,
complementing the markdown/llms.txt AEO work.
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
SEO/AEO improvements for the docs site:
- Append `| Chat SDK` to every page's `<title>` and `og:title` (home
stays `Chat SDK`).
- Expose a plain-markdown version of every docs and adapter page via an
`sr-only` AI/LLM hint link and a `text/markdown` alternate; adapter
pages get a new `/adapters/<group>/<slug>.md` endpoint.
- Turn `llms.txt` into a sitemap-style index linking to each page's
markdown, and move the full-text concatenation to `llms-full.txt`. A
`<link rel="llms-txt">` is added to every page.
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Broad SEO/AEO pass across the docs site, adapter READMEs and AGENTS.md
files, and npm package metadata so Chat SDK content shows up better in
search engines, in LLM-driven package recommendations, and in
IDE/coding-agent context.
**Docs site**
- Adds a `## Resources` section to the Getting Started and AI overview
pages and to the Slack, Discord, GitHub, Liveblocks, and Sendblue
adapter pages, each linking to applicable guides/templates with
descriptions sourced from `resources-edge-config.json` and a cross-link
back to the central `/resources` hub.
**Adapter packages**
- Mirrors the same Resources sections into the Slack, Discord, and
GitHub READMEs (so they surface on npm) and into their AGENTS.md files
(so coding agents see them alongside the API notes).
- Expands `keywords` on every published adapter and state package — adds
`chat-sdk`, `chatbot`, `ai-agent`, `ai-sdk`, `vercel`, plus
platform-specific terms like `slack-bot`, `block-kit`, `slash-commands`,
`github-app`, `whatsapp-business`, `state-adapter`.
**Resources registry**
- Registers four new entries in `resources-edge-config.json`
(Human-in-the-Loop guide, Liveblocks AI agent guide, Slack + Vercel Blob
guide, Durable iMessage Agent template) and runs `pnpm sync-resources`
so the bundled `chat` package guides, `templates.json`, and
`skills/chat/SKILL.md` all pick them up.
- Fixes the synced Slack AI agent guide to import `toAiMessages` from
`chat/ai` instead of the deprecated `chat` re-export path (the upstream
KB source has also been updated, so future syncs will preserve this).
**Drive-by fixes**
- Resend adapter doc quick start: corrects `MemoryStateAdapter` class
import to the `createMemoryState()` factory (matching every other
adapter doc).
- Zalo adapter doc: drops the "community adapter" callout that
duplicated frontmatter.
**Tooling / CI**
- Adds `tsx` as a root devDependency so `pnpm sync-resources` works out
of the box (it previously relied on `npx tsx`, which hung when not
pre-cached).
- Loosens the CI changeset gate to also skip `packages/chat/resources/`
(generated data), matching the existing `*.md` carve-out.
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
## summary
adds a first-class Twilio adapter for SMS and MMS bots, plus low-level
voice helpers for custom Twilio voice routes
this includes webhook parsing and signature verification, outbound
Messages API helpers, phone-number and Messaging Service sending,
inbound MMS attachments with authenticated `fetchData`, plain text card
fallback rendering, markdown conversion, and runtime-light `api`,
`webhook`, `voice`, and `format` subpaths
the adapter intentionally avoids the `twilio` npm runtime dependency so
apps can use the low-level helpers without pulling in the full SDK
## Summary
Refreshes the adapter docs end-to-end so every adapter — official,
vendor-official, and community — now ships hand-authored MDX, lives
under a clean URL structure, and renders on a polished
sidebar/right-rail layout dedicated to `/adapters` (the shared `/docs`
chrome is untouched).
```mermaid
flowchart LR
subgraph Before
direction TB
OB[official] --> CB[community<br/>incl. 5 vendor pages]
end
subgraph After
direction TB
OA[official] --> VA[vendor-official<br/>5 pages] --> CA[community]
end
Before -.-> After
```
### Content & routing
- **New `/adapters/vendor-official/<slug>` route** for vendor-maintained
adapters (Beeper Matrix, Photon iMessage, Liveblocks, Resend, Zernio).
Sidebar gets a third labelled group ("Vendor-Official Adapters") between
Official and Community, with a top divider matching the existing
Community treatment.
- **All 13 vendor-official + community adapters migrated** from runtime
README fetching to hand-authored MDX with rich `features:` matrices and
full body content (install, quick start, configuration, auth,
gateway/streaming, troubleshooting). README fetch stays as a fallback
for any future community adapter that hasn't been migrated yet, gated by
a new `mdxBody: true` frontmatter flag.
- **Messenger filter pages removed** (`/adapters/for/<messenger>` + the
"Browse by messenger" chip row on `/adapters`). Existing URLs
308-redirect to `/adapters`.
- **Permanent redirects** from
`/adapters/community/{matrix,imessage,resend,zernio,liveblocks}` to
their new `/adapters/vendor-official/...` paths.
- **Fixed** `/docs/adapters` and `/docs/state` so the bare pages are
accessible again — the previous catch-all redirect (`:slug*`) was
swallowing them. Switched to `:slug+` so subpath URLs still 308 while
the bare pages render.
### Visual polish
- **Adapter-only sidebar variant** (`AdaptersDocsLayout` +
`AdaptersSidebar`) with uppercase eyebrow separators, tighter rows, and
a thin themed scrollbar utility class. The shared `/docs` sidebar is
untouched.
- **Restyled `AdapterHero`**: drops the badges row + packageName, sits
the title inline with the logo, larger 17 px tagline, horizontal divider
beneath the block.
- **Restyled `PackageInstall`** as a tabbed dark single-line snippet
with a `$` prompt prefix and a copy button — replaces the previous
multi-line `CodeBlock` layout.
- **New "Deploy your chat app on Vercel" upsell card** (`<Upsell />`)
replaces the old `EditSource / ScrollTop / Feedback / CopyPage` footer
cluster on every adapter detail page.
- **Listing & messenger pages**: align the H1 to a tighter `text-4xl
sm:text-[44px]`, and the section headers to `text-base font-medium
tracking-tight` with a one-line muted lede.
### Tooling & tests
- Added `mdxBody: true` opt-in to the adapter frontmatter schema
(`source.config.ts`), and updated both detail-page handlers
(`community/[slug]` and the new `vendor-official/[slug]`) to render the
MDX body when present, falling back to README fetch otherwise.
- Refactored both detail-page handlers to flatten the body-render
branches into a `renderBody()` helper, removing the nested ternaries
that were tripping `lint/style/noNestedTernary`.
- New test file
[`packages/integration-tests/src/docs-adapters.test.ts`](https://github.com/vercel/chat/blob/docs/refresh-adapters/packages/integration-tests/src/docs-adapters.test.ts)
— **220 new assertions** covering:
- Adapter MDX frontmatter completeness, slug ↔ filename consistency, and
`type ∈ {platform, state}`.
- Vendor-official invariants: exactly the expected slugs,
`vendorOfficial: true`, `community: true`, `author`, `mdxBody: true`,
`<FeatureSupport />` rendered.
- Community invariants: `community: true` (never vendor-official),
`mdxBody: true`, `<FeatureSupport />`.
- Official invariants: never flagged, `packageName` always under
`@chat-adapter/*`.
- `adapters.json` ↔ MDX sync on `packageName` / `type` / `community` /
`vendorOfficial`.
- Extended `VALID_DOC_PACKAGES` so `docs-content.test.ts` accepts the
new vendor-official + community packages, plus `@chat-adapter/web`,
`@chat-adapter/web/react`, and `@chat-adapter/messenger`.
### Per-package AGENTS.md
- Added `AGENTS.md` to every official adapter and state adapter (14
packages), each tailored to that adapter's surface — overview, directory
layout, build/test commands, public exports, thread ID format, webhook
flow, authentication, format conversion, cards/streaming, platform
quirks, testing approach, coding conventions, and release rules.
- Added a one-line `CLAUDE.md` (`@AGENTS.md`) beside each so Claude Code
picks up the same instructions through its built-in resolver — same
convention as the root.
### Web adapter copy
- Cleaned up the Web adapter tagline (removed inline backticks) and
dropped the now-redundant "v1 scope" section from the body.
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
* chore: Add Messenger logo to homepage
* chore: Use Messenger brand colour for logo
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
* [messenger] add messenger (meta) platform adapter to chat sdk
- Webhook handling with HMAC-SHA256 signature verification
- Generic and Button template support for cards
- Postback, reaction, delivery/read confirmation handling
- Message caching for fetchMessages (Messenger has no history API)
- Replay tests and ~98% code coverage
Co-authored-by: Dimitar K. Nikolov <mitkodkn@users.noreply.github.com>
Co-authored-by: Ben Sabic <27636870+bensabic@users.noreply.github.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
* Fix: The `@chat-adapter/messenger` package version is `4.15.0` while all other packages in the Changesets fixed version group are at `4.27.0`, breaking the fixed versioning contract.
This commit fixes the issue reported at packages/adapter-messenger/package.json:3
**Bug explanation:**
The repository uses Changesets with a `"fixed"` configuration: `[["chat", "@chat-adapter/*"]]`. This means all packages matching these patterns must always share the same version number. Every package in the group (`chat`, `@chat-adapter/discord`, `@chat-adapter/gchat`, `@chat-adapter/github`, `@chat-adapter/linear`, `@chat-adapter/shared`, `@chat-adapter/slack`, `@chat-adapter/teams`, `@chat-adapter/telegram`, `@chat-adapter/web`, `@chat-adapter/whatsapp`, and the state packages) is at version `4.27.0`, except `@chat-adapter/messenger` which is at `4.15.0`.
This is likely because the messenger adapter was newly added to the monorepo (copied from a template or created fresh) and its version was never aligned with the rest of the fixed group. This mismatch will cause problems with the Changesets release workflow — when Changesets tries to bump versions for the fixed group, it may produce inconsistent or errored releases because one package is 12 minor versions behind the others.
**Fix explanation:**
Changed `"version": "4.15.0"` to `"version": "4.27.0"` in `packages/adapter-messenger/package.json` to align it with all other packages in the fixed version group.
Co-authored-by: Vercel <vercel[bot]@users.noreply.github.com>
Co-authored-by: visyat <vishal.yathish@gmail.com>
* Fix: Messenger adapter env var guard only checks `FACEBOOK_APP_SECRET` but `createMessengerAdapter` requires all three env vars, causing a `ValidationError` crash at Next.js build time when only `FACEBOOK_APP_SECRET` is set.
This commit fixes the issue reported at examples/nextjs-chat/src/lib/adapters.ts:154
**Bug Analysis:**
The build failure is confirmed in the Vercel build log with:
```
Error [ValidationError]: pageAccessToken is required. Set FACEBOOK_PAGE_ACCESS_TOKEN or provide it in config.
```
The root cause is in `examples/nextjs-chat/src/lib/adapters.ts` at line ~154. The messenger adapter guard only checks for `FACEBOOK_APP_SECRET`:
```typescript
if (process.env.FACEBOOK_APP_SECRET) {
```
However, `createMessengerAdapter` (in `packages/adapter-messenger/src/index.ts`) validates and throws `ValidationError` for each of three required env vars: `FACEBOOK_APP_SECRET`, `FACEBOOK_PAGE_ACCESS_TOKEN`, and `FACEBOOK_VERIFY_TOKEN`. When only `FACEBOOK_APP_SECRET` is set in the Vercel project environment, the guard passes, `createMessengerAdapter` is called, and it throws a `ValidationError` for the missing `FACEBOOK_PAGE_ACCESS_TOKEN`. Since this code runs at module evaluation time during the Next.js build's "Collecting page data" phase, the uncaught error crashes the entire build.
This is inconsistent with other adapters in the same file. For example, the WhatsApp adapter checks both `WHATSAPP_ACCESS_TOKEN` and `WHATSAPP_PHONE_NUMBER_ID`, and the gchat/github/linear/whatsapp adapters all wrap creation in try-catch blocks.
**Fix:**
1. Updated the env var guard to check all three required environment variables (`FACEBOOK_APP_SECRET`, `FACEBOOK_PAGE_ACCESS_TOKEN`, and `FACEBOOK_VERIFY_TOKEN`) before attempting to create the adapter.
2. Wrapped the `createMessengerAdapter` call in a try-catch block (matching the pattern used by gchat, github, linear, and whatsapp adapters) so that any unexpected validation errors are caught and logged as warnings instead of crashing the build.
Co-authored-by: Vercel <vercel[bot]@users.noreply.github.com>
Co-authored-by: visyat <vishal.yathish@gmail.com>
---------
Co-authored-by: Dimitar K. Nikolov <mitkodkn@users.noreply.github.com>
Co-authored-by: Ben Sabic <27636870+bensabic@users.noreply.github.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: vercel[bot] <35613825+vercel[bot]@users.noreply.github.com>
Co-authored-by: Vercel <vercel[bot]@users.noreply.github.com>
* feat(chat): expose awaitable Promise from processMessage
Return the inner task as Promise<void> instead of void so streaming
adapters can await full handler completion and surface user-handler
rejections at the wire level. waitUntil semantics for existing webhook
adapters are unchanged — the SDK still tracks the work with errors
swallowed (and logged) so platforms don't retry on handler bugs.
Required by @chat-adapter/web, whose response body is the user
handler's stream.
* feat(adapter-web): add @chat-adapter/web package
A new platform adapter that lets a chat-sdk bot serve a browser chat
UI alongside Slack/Teams/Discord/etc. without writing any client-side
glue. Speaks the AI SDK UI message stream protocol, so @ai-sdk/react's
useChat and the ai-elements component library work out of the box.
- `@chat-adapter/web` — server: createWebAdapter({ userName, getUser })
- `@chat-adapter/web/react` — client: useChat() preconfigured with
DefaultChatTransport against /api/chat (override via `api`)
Defaults that matter for v1:
- `isDM: true` — every web message routes through onDirectMessage
- `persistMessageHistory: true` — chat-sdk caches each turn in the
configured state adapter so handlers can read prior context via
thread.messages / channel.messages (no platform history API exists)
- channelId === threadId — web has no separate channel concept; this
prevents cross-conversation bleed when a single user has multiple
useChat sessions
- Native `adapter.stream` implementation pumps text-deltas straight
onto the SSE response — no post+edit fallback
Out of scope for v1: cards/JSX rendering, reactions, modals, file
uploads, edit/delete, multi-tab proactive push.
* feat(example-nextjs-chat): wire up web adapter and add /chat page
- Register the web adapter in lib/adapters.ts with a demo getUser
(single shared identity — replace with NextAuth/Clerk/cookie auth
in production)
- Expose POST /api/chat backed by bot.webhooks.web (using next/after
for waitUntil)
- Add a minimal /chat page using @chat-adapter/web/react's useChat —
same bot.onDirectMessage handler that powers Slack now powers the
browser too
Bumps `ai` to ^6.0.174 to align with @ai-sdk/react@^3 (avoids dual
provider-utils versions in the workspace).
* docs: list @chat-adapter/web in registry
- Add an entry to adapters.json so the package shows up on /adapters
- Add a globe SVG to lib/logos.tsx and wire it into the icon map
- Mention the new adapter in docs/adapters.mdx
* feat(adapter-web): tighten request handling and message construction
- Reject user ids containing ':' with HTTP 400 — the character would
corrupt the thread-id round-trip through decodeThreadId
- Skip emitting text-start/text-end in postMessage when the resolved
text is empty so useChat doesn't render blank assistant bubbles
- Derive the parseMessage author from raw.role so rehydrated assistant
messages report the bot identity instead of "unknown"
- Drop the duplicate handler-error log; chat.processMessage already
logs at ERROR level
- Document the actual persistMessageHistory default (true) and the
state-cache rationale; promote the fetchMessages no-op rationale
into its JSDoc
* test(adapter-web): add direct coverage for stream()
- Aborting request.signal mid-stream short-circuits the iterator and
still writes text-end via the finally block
- Non-text StreamChunks (task_update, plan_update) are dropped without
emitting any delta
- The SentMessage returned from thread.post matches the id used in
text-start / text-end events
* docs(adapter-web): expand README into the full adapter docs page
The docs site renders each adapter's README, so flesh out
@chat-adapter/web to match the depth of @chat-adapter/slack:
authentication boundary, threading semantics, streaming,
persistence, React hook reference, configuration table,
feature matrix, and troubleshooting.
* docs(adapter-web): drop unsupported provider import from streaming example
* fix(adapter-web): validate conversationId for reserved colon character
* fix(example): show error state in web chat demo
* fix(example): add thinking indicator to web chat demo
* feat(example): redesign web chat demo with tailwind
* chore: remove redundant changeset
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Co-authored-by: dancer <josh@afterima.ge>
- Add scripts/sync-resources.ts (run via pnpm sync-resources) that reads apps/docs/resources-edge-config.json, fetches each guide's .md version over https with a timeout and size cap, writes templates.json, and regenerates the Available resources block in skills/chat/SKILL.md
- Migrate the Slack Next.js, Discord Nuxt, and Hono code-review guides from on-site MDX to Vercel KB and register them in the resources edge-config JSON alongside the existing external guides
- Remove /docs/guides MDX content, sidebar entries, top-level Guides nav entry, getting-started cards, and the dead /guides/ branch in the sitemap route now that all guides live externally and are surfaced on /resources
- Replace the homepage Guides/Templates section and the standalone Adapters pill section with a single two-column Resources + Adapters section (icons, headings, descriptions, outline buttons, divider), and drop the URL footer from ResourceCard on the Resources page
- Update skills/chat/SKILL.md to point at resources/guides and resources/templates.json and list the available guides and templates between marker comments that sync-resources rewrites
- Add tsx to knip's ignoreBinaries so npx tsx in the new script does not fail lint
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Co-authored-by: dancer <josh@afterima.ge>
Restructures the docs homepage around the AI SDK section flow: hero
with command prompt + interactive demo, OSS stats grid, supported
platforms with feature pillars, Chat SDK Core code showcase, "Scale
with confidence" Vercel ecosystem integrations, and a closing
get-started section with three guides.
- New: oss-stats-section, supported-platforms, code-section,
integrations-section, get-started-section.
- Removed unused layout helpers: adapters-section, centered-section,
cta, one-two-section, text-grid-section, templates (folded into
get-started-section).
- Hero loses the badge and embeds the existing Demo as the
hero-interactive piece.
- Three guide cards link to the Slack agent guide, Knowledge Agent
template, and GitHub code review bot guide.
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
* docs: sync geistdocs template 1.2.0–1.2.3 + polish
- Ran @vercel/geistdocs@1.2.2 update --sync against origin/main
(already contains merged 1.2.3) for:
- components/geistdocs/*
- components/geistcn-fallbacks/**/* (new)
- components/ui/command-prompt.tsx, navigation-menu.tsx
- app/styles/geistdocs.css
- Manual overlays (paths skipped by sync due to chat customizations):
- app/[lang]/layout.tsx: drop scroll-smooth (1.2.2)
- app/[lang]/docs/[[...slug]]/page.tsx: MobileDocsBar + disable
default TOC popover (1.2.0)
- app/[lang]/docs/layout.tsx: wrap in bg-background-200 (1.2.3)
- components/ui/badge.tsx: secondary variant → bg-gray-300/text-gray-1000 (1.2.3)
- Home hero: replace Get Started + Installer with CommandPrompt
humans/agents switcher ("npm install chat" / "npx skills add vercel/chat")
- (home) layout: swap bg-sidebar dark:bg-background for bg-background-200
so /, /adapters, /resources share the navbar surface
- DesktopMenu: active-state detection with longest-prefix match
(so /docs/api highlights "API", not also "Docs")
- navbar-logo dropdown: drop Chat SDK self-entry
- New geist-fill icons (check-circle-fill, cross-circle-fill,
warning-fill) ported from @vercel/geistcn-assets; new
components/custom/status-icons.tsx registers Check/Cross/Warn MDX
components
- content/docs/adapters.mdx: replace ✅/❌/⚠️ emojis with the new
icons (emoji.mdx intentionally left alone)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* docs: use LogoChatSdk from geistcn-fallbacks as the app Logo
Replace the inline Chat SDK wordmark SVG in geistdocs.tsx with
<LogoChatSdk /> so the navbar and other Logo consumers share the
same source of truth.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* docs: default hero CommandPrompt to humans tab
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* feat(docs): add Resources page powered by Edge Config
Add a /resources page that displays guides and templates in a
3-column card grid. Data is fetched from Vercel Edge Config in
production and from a local JSON file in development. Includes
CollectionPage JSON-LD markup and revalidates daily.
* fix(docs): handle Edge Config unavailability on resources page
* feat(docs): add keywords metadata to resources page
* feat(docs): add twitter card metadata to resources page
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
* docs: add agent readability improvements
- Add AGENTS.md skill file served at /AGENTS.md
- Add /.well-known/agent-skills/ discovery index and SKILL.md
- Add JSON-LD structured data to homepage
- Exclude AGENTS.md and .well-known from i18n middleware
* fix: suppress biome lint for JSON-LD dangerouslySetInnerHTML
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
* docs: misc updates and fixes
* docs: update feature matrix to show Teams modal support
---------
Co-authored-by: Ben Sabic <bensabic@users.noreply.github.com>
Rich Haines