Sign in Book a demo
Back to Tools
Developer Tools API Key
Build Your Agent View Vercel Docs
vercel logo

Vercel MCP Tools for AI Agents

Let an agent create projects, trigger deployments, read build logs, manage domains and environment variables, and roll production back.

Available Tools

25 tools available in this integration

Vercel.AddProjectDomain

Attach a custom domain to a project; the result reports its verification status.

Vercel.CancelDeployment

Cancel an in-progress deployment.

Vercel.CreateDeployment

Trigger a new deployment from inline source files, a git repository, or a prior deployment.

Vercel.CreateProject

Create a new project, optionally connecting a git repository to deploy from.

Vercel.DeleteEnvironmentVariable

Delete an environment variable from a project.

Vercel.DeleteProject

Permanently delete a project and all of its deployments.

Vercel.GetCurrentUser

Return the Vercel account the access token authenticates as.

Vercel.GetDeployment

Return a single deployment's state and metadata.

Vercel.GetDeploymentLogs

Read a deployment's build and runtime log events to diagnose why it failed.

Vercel.GetProject

Return a single project's settings and latest deployment state.

Vercel.GetProjectDomain

Return a project domain's verification status and DNS-configuration state.

Vercel.GetTeam

Return a team's details, including its billing plan.

Vercel.ListDeploymentLogs

Read a deployment's build and runtime log events to diagnose why it failed.

Vercel.ListDeployments

List deployments in the active scope, optionally filtered by project, target, and state.

Vercel.ListEnvironmentVariables

List a project's environment variables across all environments.

Vercel.ListProjectDomains

List the custom domains attached to a project.

Vercel.ListProjects

List the projects in the active scope, optionally narrowed by a fuzzy name search.

Vercel.ListTeams

List the teams the access token can act on, to discover a valid team scope.

Vercel.PromoteDeployment

Make a deployment the project's current production deployment ("ship this one").

Vercel.RemoveProjectDomain

Detach a custom domain from a project.

Vercel.RollbackDeployment

Roll production back to a prior deployment, making it the current production deployment.

Vercel.SaveEnvironmentVariable

Create or update a project environment variable.

Vercel.UpdateProjectSettings

Update a project's build and framework settings; unset fields are left unchanged.

Vercel.VerifyProjectDomain

Trigger verification of a project domain and return its resulting status.

Vercel.WhoAmI

Return the Vercel account the access token authenticates as.

Working directly against the raw API? 194 more tools in the Vercel API server →

Why agent-optimized tools matter

Built for intent, not API schemas

Ask the agent to "find the latest failed deployment for my project, explain the build error, and tell me what to change." Arcade picks the right Vercel actions without guessing project or deployment identifiers.

Up to 100x fewer tokens

Vercel list and search results return project, deployment, domain, and variable metadata before heavier logs or settings are fetched, so the agent can narrow the task without loading every deployment record into context.

Complete the workflow end to end

Inspect the project, read deployment logs, update settings, create a deployment, and promote the release through one toolkit designed around the complete Vercel deployment task.

Make actions predictable and controlled

Arcade tools return structured results and useful errors across users and runs. Teams can apply scope restrictions, redaction, and approval policies so agents follow the same controls every time.

Securing Vercel

The more capabilities agents get, the more useful they are. However, more capabilities bring more risks of undesired effects such as data loss and exfiltration.

This is why Arcade provides you with contextual rules that dictate how and when a tool is allowed to run. Arcade’s contextual access lets you govern tool visibility and behavior at every stage of execution and control who can see a tool, what inputs are allowed, and what comes back.

Learn about Contextual Access Policies

For Vercel, you can implement some of these policies:

Access when tools are listed for a user
  • Hide Vercel.DeleteProject from agents unless the roles your IDP marks as platform owners are active, so an agent cannot permanently erase a project and its deployments.
  • Restrict Vercel.DeleteEnvironmentVariable to agents operating under your environment-administrator role, so routine automation cannot remove variables that builds depend on.
  • Deny Vercel.RemoveProjectDomain to everyone outside the roles your IDP marks as domain administrators, so an agent cannot take a production site offline by detaching its domain.
Pre-execution before each tool call
  • Require human approval before Vercel.PromoteDeployment runs, so an agent cannot make an unreviewed build the live production deployment.
  • Block Vercel.SaveEnvironmentVariable when the variable name targets production credentials or signing keys without an approved change ticket, so an agent cannot replace secrets used by live workloads.
  • Require human approval before Vercel.CreateDeployment runs against a production target, so an agent cannot publish unreviewed source code.
Post-execution after the tool returns
  • Redact environment-variable values and any secret-looking metadata from Vercel.ListEnvironmentVariables before they reach the model, so credentials cannot enter its context.
  • Strip access tokens, API keys, connection strings, and customer records from Vercel.GetDeploymentLogs results before the model sees build or runtime output.
  • Deny the output of Vercel.GetCurrentUser when it contains personal email addresses or account identifiers not needed for the requested task, and record the refusal for your audit trail.

How to use Vercel with your agent

Connect Vercel to an MCP client through Arcade, then run deployment and project workflows from prompts. Follow the walkthrough to configure the shared API key and test the connection.

Your questions, answered

A Vercel MCP server exposes Vercel capabilities as structured tools that an AI agent can call through the Model Context Protocol. With a configured Vercel credential, the agent can inspect projects and deployments, read logs, manage domains and environment variables, and control releases within that credential's access.

Vercel does not provide the MCP server used here; Arcade hosts the integration and adds the runtime around Vercel access. That includes credential brokering, approval policies, response redaction, tool visibility controls, and one MCP or tool API connection for supported clients.

Arcade agents can list and inspect projects, teams, deployments, domains, and environment variables; read build and runtime logs; create or cancel deployments; create projects; update project settings; add, verify, or remove domains; save or delete environment variables; promote or roll back production deployments; and delete projects. Additional operations are available through the Vercel API surface.

This integration uses a shared Vercel API key or secret configured once. The actions available to agents are limited by the permissions and scope granted to that credential, so create, deployment, environment-variable, domain, promotion, rollback, and deletion actions should use a key with only the access your workflow needs.

No. Users do not each authorize a separate Vercel account through this integration; one shared API key or secret is configured for the connection. Every agent call runs with the access granted to that credential, so separate user-level Vercel boundaries require separate connections or credentials.

Yes. Contextual Access policies can require human approval before consequential Vercel actions such as promoting, rolling back, canceling, or deleting a deployment, changing environment variables, removing a domain, or deleting a project. The call proceeds only after the policy's conditions have been satisfied.

Arcade proxies Vercel API calls and does not persist your content as part of normal tool execution. The Vercel credential is brokered and stored encrypted at rest, and the model never receives it.

Any framework or client that can connect to Arcade's MCP gateway or tool APIs works, including LangChain, OpenAI Agents, CrewAI, Claude Desktop, Claude Code, Cursor, Visual Studio Code, Microsoft Copilot Studio, and GitHub Copilot.

Arcade refreshes access while the configured Vercel credential remains valid. If the key is revoked, rotated, expired, or can no longer be used, calls stop until the credential is updated or reconnected.

Calls remain subject to Vercel API rate limits and quotas for the account or team behind the configured credential. Arcade does not remove those Vercel limits, so high-volume deployment, project, log, or environment-variable workflows should handle throttling and retries.

Ready to actually ship your agent?

Connect your agent to Vercel with a shared API key, agent-optimized tools and approval controls for consequential actions.

Build Your Agent Read the Vercel Quickstart