Vercel Connect logo

Vercel Connect

Organization
vercel
vercel-connect

Vercel Connect expert guidance — securely obtain scoped OAuth tokens for third-party services (Slack, GitHub, MCP servers, OAuth, Snowflake) on behalf of apps or users via Vercel OIDC. Use when wiring up third-party API access, connecting to MCP servers, sending Slack messages, accessing GitHub APIs, receiving webhook events from Slack/Linear/GitHub and forwarding them to your agents and apps, or building eve agent connections.

Overview

Publishervercel
Repositoryvercel-plugin
Skill namevercel-connect
Stars
286
Forks
56
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by vercel on GitHub. Read the source before you install it.

Installation

Install the Vercel Connect AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/vercel/vercel-plugin.git /tmp/vercel-plugin
mkdir -p .claude/skills
cp -r /tmp/vercel-plugin/skills/vercel-connect .claude/skills/vercel-connect
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Vercel Connect in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Vercel Connect on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Vercel Connect is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Vercel Connect Skill

Overview

Vercel Connect enables to securely obtain scoped tokens for accessing third-party services on behalf of apps or users. It uses Vercel OIDC tokens to authenticate and exchange for Vercel Connect tokens via the Vercel API.

When to Use Vercel Connect

Use Vercel Connect when you need to:

  • Send messages via Slack (as a bot or on behalf of a user)
  • Access GitHub repositories or APIs
  • Connect to any third-party system that requires OAuth tokens or API credentials
  • Obtain tokens for authenticated API calls

Modes of tokens

The SDK supports three subject types — pick based on what's acting:

  • user — actions performed on behalf of a specific end user (e.g., post a Slack message as the user). Requires a user id and optional issuer.
  • app — actions performed as the app itself (e.g., post as a Slack bot, app-level GitHub access). No consent flow — fails terminally if the connector is not installed.
  • jwt-bearer — RFC 7523 JWT-bearer exchange for connectors that accept a caller-minted assertion. Pass sub (required), plus optional iss, aud, and additionalClaims. Use when the third-party expects you to vouch for the subject via a signed JWT rather than an interactive OAuth grant.

Available Tools

All tools have --format=json option for machine-readable output.

1. Vercel Connect CLI (for Bash/Shell)

Use the vercel connect CLI for command-line operations. Use vercel connect --help to get available commands. The user needs to be authenticated to the Vercel CLI and the commands work within the scope of the user's currently selected Vercel team. For eg it will create & list Connect connectors created within the currently selected Vercel team.

Important! Always run vercel connect commands from the project or agent folder that will consume the connection (the directory containing package.json / vercel.json). Vercel Connect reads the local project context to auto-configure the connection — for example, picking a sensible connector name and uid, setting up project access to the connection, configuring webhooks and triggers. Running from the repo root or an unrelated directory skips this auto-configuration and you'll have to wire things up by hand. If the user invokes a vc connect command from elsewhere, cd into the closest matching project/agent folder first (or pass --cwd <DIR>).

Example commands:

bash
# Create new Connect connector
vercel connect create <service>

# List existing Connect connectors
vercel connect list

# Get token
vercel connect token <connector> --subject user|app

Important! The vercel connect create and vercel connect token commands may open the browser for the user if there's a manual registration required (for eg completing the OAuth consent or installing a slack app to a workspace). The user must visit the browser to complete the process while you wait for the process to complete.

Available Services
ServiceModesDescription
slackuser, botSlack API access
githubuser, appGitHub API access
MCP serversuser, appAny MCP server (mcp.<host>/<path>)
snowflakeuserSnowflake data access
Generic OAuth provideruser, appAny OAuth 2.0 server registered via vercel connect create

For MCP servers, pass the full endpoint URL when registering (e.g. vercel connect create https://mcp.linear.app/mcp). The connector ID then takes the form mcp.<host>/<name> (for example mcp.linear.app/myagent).

Example: Send a Slack message using curl
bash
TOKEN=$(vercel connect token <connector>)
curl -X POST https://slack.com/api/chat.postMessage \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"channel": "C1234567890", "text": "Hello from Vercel Connect!"}'

2. JavaScript/TypeScript SDK (@vercel/connect)

For JavaScript/TypeScript code, use the @vercel/connect package directly:

typescript
import { getToken } from "@vercel/connect";

// Get a token for Slack bot
const token = await getToken("scl_abc123", {
  subject: { type: "app" }, // If sending as a bot, or else use "user"
});

// Use the token
const response = await fetch("https://slack.com/api/chat.postMessage", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${token}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    channel: "C1234567890",
    text: "Hello from Vercel Connect!",
  }),
});

The SDK uses the user's Vercel OIDC token to authenticate. The user should have run vc env pull to pull the OIDC token env variables locally (or vc link pulls it automatically)

eve agents — @vercel/connect/eve

When the project is built on eve, prefer the connect helper over calling getToken directly inside connection definitions. The helper wires the full token / start-authorization / complete-authorization lifecycle into eve's connection runtime, so a Vercel Connect-backed connection becomes a single declaration:

typescript
// agent/connections/linear.ts
import { defineMcpClientConnection } from "eve/connections";
import { connect } from "@vercel/connect/eve";

export default defineMcpClientConnection({
  url: "https://mcp.linear.app/mcp",
  description: "Linear workspace — issues, projects, cycles, and comments.",
  auth: connect("mcp.linear.app/myagent"),
});

Key points for the agent:

  • Omit principalType for the default per-user OAuth flow, or set "app" for app-scoped tokens (no consent flow — fail terminally if not installed).
  • Pass the connector id directly with connect("mcp.linear.app/myagent"), or use connect({ connector: "mcp.linear.app/myagent" }) when you need options.
  • For scopes, audiences, or authorizationDetails, pass them through tokenParams. For a custom challenge prompt, pass instructions. Both are optional.
  • eve is an optional peer dependency, so the rest of @vercel/connect (CLI, getToken, etc.) is unaffected for non-eve consumers.
Slack channel — connectSlackCredentials

For eve Slack channels (agent/channels/slack.ts), use connectSlackCredentials(connector) from @vercel/connect/eve. It returns a complete SlackChannelCredentials object — both the bot token and inbound webhook verification are handled by Vercel Connect, so you do not need SLACK_BOT_TOKEN or SLACK_SIGNING_SECRET env vars:

typescript
// agent/channels/slack.ts
import { slackRoute } from "eve/channels/slack";
import { connectSlackCredentials } from "@vercel/connect/eve";

export default slackRoute({
  credentials: connectSlackCredentials("slack/myagent"),
});

What the helper wires up:

  • botToken: a function that calls getToken(connector, { subject: { type: "app" } }) on each inbound webhook, so token rotation, refresh, and multi-workspace tenancy are handled server-side.
  • webhookVerifier: a Vercel OIDC verifier (vercelOidc()). Vercel Connect forwards verified Slack webhooks to your app as signed Vercel OIDC requests; the helper verifies that signature instead of the raw Slack signing secret.

Use this whenever the project is on eve + Vercel Connect — it's the one-liner for both outbound posts and inbound webhook auth.

GitHub channel — connectGitHubCredentials

For eve GitHub channels (agent/channels/github.ts), use connectGitHubCredentials(connector) from @vercel/connect/eve. It returns a complete GitHubChannelCredentials object — eve uses the installation token directly (skipping its native GitHub App JWT exchange) and Vercel Connect handles rotation, refresh, and multi-installation tenancy server-side. You do not need GITHUB_APP_PRIVATE_KEY, GITHUB_APP_ID, GITHUB_INSTALLATION_ID, or GITHUB_WEBHOOK_SECRET env vars:

typescript
// agent/channels/github.ts
import { githubRoute } from "eve/channels/github";
import { connectGitHubCredentials } from "@vercel/connect/eve";

export default githubRoute({
  credentials: connectGitHubCredentials("github/myagent"),
});

What the helper wires up:

  • installationToken: a function that calls getToken(connector, { subject: { type: "app" } }). The helper pins subject to "app" — GitHub installation tokens are app-scoped.
  • webhookVerifier: a Vercel OIDC verifier (vercelOidc()). Vercel Connect forwards verified GitHub webhooks to your app as signed Vercel OIDC requests; the helper verifies that signature instead of the raw GitHub webhook secret.
Linear channel — connectLinearCredentials

For eve Linear channels (agent/channels/linear.ts), use connectLinearCredentials(connector) from @vercel/connect/eve. It returns a complete LinearChannelCredentials object — Vercel Connect manages the Linear app access token and webhook auth, so you do not need LINEAR_API_KEY or LINEAR_WEBHOOK_SECRET env vars:

typescript
// agent/channels/linear.ts
import { linearRoute } from "eve/channels/linear";
import { connectLinearCredentials } from "@vercel/connect/eve";

export default linearRoute({
  credentials: connectLinearCredentials("linear/myagent"),
});

What the helper wires up:

  • accessToken: a function that calls getToken(connector, { subject: { type: "app" } }). The helper pins subject to "app" — Linear Agent tokens are app-scoped.
  • webhookVerifier: a Vercel OIDC verifier (vercelOidc()). Vercel Connect forwards verified Linear webhooks to your app as signed Vercel OIDC requests; the helper verifies that signature instead of the raw Linear webhook secret.

3. HTTP API (for other languages)

For other languages, make HTTP requests directly to the Vercel Connect server. The request must be authenticated with the project's Vercel OIDC token (VERCEL_OIDC_TOKEN env var — pulled by vc env pull or injected at runtime):

bash
# Get a token via HTTP
POST https://api.vercel.com/v1/connect/token/<connector>
Authorization: Bearer <VERCEL_OIDC_TOKEN>
Content-Type: application/json

{ "subject": { "type": "user", "id": "user_123" } }

The response is JSON with a token field (plus expiresAt, connector, and other metadata).

Python Example
python
import os
import requests

# Get token from Vercel Connect
connect_response = requests.post(
    "https://api.vercel.com/v1/connect/token/slack1234",
    headers={"Authorization": f"Bearer {os.environ['VERCEL_OIDC_TOKEN']}"},
    json={
        "subject": {"type": "app"},
    },
)
token = connect_response.json()["token"]

# Use the token
slack_response = requests.post(
    "https://slack.com/api/chat.postMessage",
    headers={"Authorization": f"Bearer {token}"},
    json={"channel": "C1234567890", "text": "Hello from Vercel Connect!"}
)

4. BetterAuth and AuthJS support

When the app already uses Better Auth or Auth.js for end-user authentication, you can plug a Vercel Connect connector in as an OAuth provider instead of calling getToken directly. The connect helper on each subpath handles the token exchange so provider credentials stay in Vercel Connect rather than in framework config or env vars.

Better Auth — @vercel/connect/betterauth

Optional peer dependency: better-auth. Pass the connector through Better Auth's genericOAuth plugin. Connector UIDs can contain a / (e.g. linear/myagent), and Better Auth additionally requires a providerId:

typescript
import { genericOAuth } from "better-auth/plugins";
import { connect } from "@vercel/connect/betterauth";

genericOAuth({
  config: [connect({ providerId: "linear", connector: "linear/myagent" })],
});
Auth.js — @vercel/connect/authjs

Optional peer dependency: @auth/core. Use the connector as an OAuth2Config provider. Connector UIDs can contain a / (e.g. linear/myagent), and Auth.js additionally requires an id:

typescript
import { connect } from "@vercel/connect/authjs";

const providers = [connect({ id: "linear", connector: "linear/myagent" })];

Workflow

All tools have --json option for machine-readable output.

Before running any vercel connect step below, make sure your shell cwd is the project or agent folder that will use the connection (see the CLI section above). Vercel Connect uses that context to auto-configure the project, so running from the right directory removes follow-up wiring work.

  1. Check existing Connect connectors: See if a required Connect connector is already present

    bash
    vercel connect list
    vercel connect token <connector>

Important! If more than one connector found, allow user to make the choice between them, or ask to create a new one

  1. Register: If the provider you need is not registered of if the user asked to create a new connector / app / bot, follow the instructions to register it (this may involve setting up credentials on browser in the third-party service and then registering them with Vercel Connect).

    bash
    vercel connect create <service> [--name <app-name>]

Important! Provide the most precise server URL for the service, including the complete connection URL (e.g. https://mcp.linear.app/mcp rather than just linear). Short service aliases may resolve to a default endpoint that does not match the transport or path the user actually wants. When in doubt, run vercel connect create --help to confirm which service names and URL forms are accepted before picking one.

Important! This command will give you a URL or directly open it to complete the registration process. User must visit that URL and follow the instructions to link their third-party account with Vercel Connect. The command will not complete until they finish the registration. The agent must clearly show the URL to the user and prompt them to complete the registration.

Important! Once vercel connect create completes, it will print a successful message. You must capture that connector ID for the next step.

Important! The vercel connect create command may open the browser so it's better to get the user approval before running it.

  1. Get token: Obtain a token for the provider you need: On CLI, you can get the token via

    bash
    vercel connect token <connector> [--subject <subject>]

The default subject is user. Use app for getting app scoped tokens. It's recommended to run this command with the --yes in case an re-authorization or installation is required. This will trigger the reauthorization flow for the user.

Important! Always put the token value into a variable and use the variable in the subsequent commands to avoid accidentally echoing the token in the terminal or logs. Avoid combining this command with other commands using &&. For example:

bash
TOKEN=$(vercel connect token)

Important! Try to reuse tokens as much as possible. If you already have a token with the required scopes, use it instead of requesting a new one, even when fewer scopes are needed. This will reduce friction for the user and avoid unnecessary authorization prompts.

When working with a JavaScript/TypeScript code, use the @vercel/connect package directly:

  1. Use token: Use the token to authenticate with the third-party service. For example:
typescript
import { getToken } from "@vercel/connect";

const token = await getToken(
  "connector-id",
  // Optional params:
  {
    subject: { ... },
  },
);

Frequently asked questions

What does the Vercel Connect AI skill do?

Vercel Connect expert guidance — securely obtain scoped OAuth tokens for third-party services (Slack, GitHub, MCP servers, OAuth, Snowflake) on behalf of apps or users via Vercel OIDC. Use when wiring up third-party API access, connecting to MCP servers, sending Slack messages, accessing GitHub APIs, receiving webhook events from Slack/Linear/GitHub and forwarding them to your agents and apps, or building eve agent connections.

Why use Vercel Connect on TypingMind?

Because you install it once and use it with any model. Vercel Connect is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Vercel Connect in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/vercel/vercel-plugin/tree/main/skills/vercel-connect. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Vercel Connect?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Vercel Connect?

As many as you like. As long as a model supports skills, you can use Vercel Connect with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Vercel Connect AI skill free?

It is published on GitHub by vercel. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇