extractCard(message) - extract a CardElement from an AdapterPostableMessage
extractFiles(message) - extract FileUpload[] from an AdapterPostableMessage
extractPostableAttachments(message) - extract Attachment[] from an AdapterPostableMessage
Buffer conversion
toBuffer(input) - convert various inputs (Buffer, Blob, string, URL) to Buffer
toBufferSync(input) - synchronous version of toBuffer
bufferToDataUri(buffer, mimeType) - convert a Buffer to a data URI string
Card conversion
cardToFallbackText(card) - render a CardElement as plain text for platforms that don't support cards
createEmojiConverter(platform) - create a function that converts emoji placeholders to platform format
mapButtonStyle(style) - map generic button styles to platform-specific variants
renderGfmTable(headers, rows) - render a GitHub Flavored Markdown table
escapeTableCell(text) - escape pipe characters in table cells
findCardActions(children) - find the first actions element, searching nested sections
truncateWithEllipsis(text, maxLength) - cut text to a length limit and end it with …
Callback data
encodeCallbackData(actionId, value?) - encode a button or quick-reply payload as chat:{json}
decodeCallbackData(data, fallbackActionId) - decode it, passing legacy and third-party payloads through unchanged
isChatCallbackData(data) - check for the chat: prefix
These are also exported from @chat-adapter/shared/callback-data, which has no dependencies, for runtime-light entry points such as webhook parsers.
Message history
For platforms whose API cannot list past messages, such as Messenger and Instagram:
upsertMessage(messages, message, compare?) - insert or replace a message by ID in a list that is already sorted, keeping it sorted
paginateMessages(messages, options) - page a sorted list with message-ID cursors
compareMessagesByTime(a, b, sequence?) and messageSequence(id) - the default sort order: send time, then the trailing :{n} in the message ID
Meta platforms
Used by @chat-adapter/messenger, @chat-adapter/instagram, and @chat-adapter/whatsapp:
verifyMetaWebhookChallenge(params, verifyToken) - check the query parameters of a webhook subscription request and return the challenge, or null
verifyMetaSignature(body, signature, appSecret) - verify the X-Hub-Signature-256 header in constant time
throwMetaGraphApiError(options) - map a failed Graph API response onto the shared error classes
HTTP headers
parseRetryAfter(header) - read a Retry-After header, given as seconds or as an HTTP date, and return the seconds to wait or undefined
Token encryption
AES-256-GCM helpers for encrypting OAuth tokens at rest before writing them to a state adapter. Used by @chat-adapter/slack and @chat-adapter/linear; available to any adapter that persists credentials.
encryptToken(plaintext, key) - encrypt a string and return an EncryptedTokenData envelope (random 12-byte IV per call)
decryptToken(data, key) - decrypt an envelope back to the original string
decodeKey(encoded) - decode a hex-64 or base64-44 encoded 32-byte key; throws on wrong length
isEncryptedTokenData(value) - type guard for distinguishing envelopes from legacy plaintext records
EncryptedTokenData - the envelope type ({ data, iv, tag }, all base64)
If you use an AI coding agent such as OpenAI Codex, Claude Code, or Cursor, install the Chat SDK skill so it knows the SDK APIs, adapter patterns, and project conventions before writing code.
npx skills add vercel/chat
The skill references bundled documentation in node_modules/chat/docs, plus adapter guides and starter templates in the published package.
You can also install the Vercel Plugin for a broader agent toolkit — it includes the Chat SDK skill alongside specialist agents, agent slash commands, and more:
npx plugins add vercel/vercel-plugin
The plugin is optional; the skill alone is enough to build with Chat SDK.