deployvera
OfficialThis server has a remotely hosted endpoint. Open its connector listing for connection details, authentication requirements, and setup instructions:
DeployVera for AI agents
Connect coding agents (Claude Code, Cursor, VS Code, Codex and any other MCP client) to your existing DeployVera projects. The agent can list projects, deploy, follow builds, read logs, get a plain-language explanation of a failed build, check managed databases, and manage environment variables. It uses DeployVera's hosted MCP server at https://mcp.deployvera.com/mcp (Streamable HTTP), listed in the official MCP Registry as com.deployvera/mcp. This repository also contains a Claude Code plugin with a deploy workflow skill, plus Cursor and Gemini CLI manifests.
Get a token (only for token-based setups)
Browser sign-in needs no token. For clients without OAuth support, or for CI, create one:
Sign in to the DeployVera panel.
Open Settings > AI agent access and create a token.
Pick the scopes (
read,deploy,env) and a lifetime (30, 90 or 365 days).Copy the token (
dvp_...). It is shown once.
Use the smallest scope set that works. A read token cannot change anything.
Related MCP server: MileHost MCP Server
Install
Sign in with browser (recommended)
Add the server URL without any token. The client opens your browser, you sign in to DeployVera and choose the scopes and duration, and the client stores the login. Nothing to copy or paste. Connected apps are listed, and can be disconnected, in Settings > AI agent access.
Claude Code: plugin (MCP server plus the deploy skill):
/plugin marketplace add deployvera/deployvera-mcp
/plugin install deployvera@deployveraThen run /mcp, select deployvera and authenticate. Or add the server only:
claude mcp add --transport http deployvera https://mcp.deployvera.com/mcpCursor:
One click adds the server; Cursor then opens the DeployVera sign-in. Or add it to ~/.cursor/mcp.json (the repository also works as a Cursor plugin with a root mcp.json):
{
"mcpServers": {
"deployvera": { "url": "https://mcp.deployvera.com/mcp" }
}
}VS Code (.vscode/mcp.json or the user mcp.json):
{
"servers": {
"deployvera": { "type": "http", "url": "https://mcp.deployvera.com/mcp" }
}
}Gemini CLI: the repository root is a Gemini CLI extension (gemini-extension.json) with the same URL, no token. Gemini CLI starts the browser login on the first 401 response.
With a token
Use this when the client has no OAuth login, or for scripts. Create a token first (see above).
Claude Code
claude mcp add --transport http deployvera https://mcp.deployvera.com/mcp --header "Authorization: Bearer dvp_..."Cursor: Add to Cursor installs a placeholder token (YOUR_DEPLOYVERA_TOKEN); replace it in Cursor's MCP settings. Or in ~/.cursor/mcp.json:
{
"mcpServers": {
"deployvera": {
"url": "https://mcp.deployvera.com/mcp",
"headers": { "Authorization": "Bearer dvp_..." }
}
}
}VS Code: VS Code asks for the token once and stores it securely.
{
"inputs": [
{ "type": "promptString", "id": "deployvera-token", "description": "DeployVera token (dvp_...)", "password": true }
],
"servers": {
"deployvera": {
"type": "http",
"url": "https://mcp.deployvera.com/mcp",
"headers": { "Authorization": "Bearer ${input:deployvera-token}" }
}
}
}Codex (~/.codex/config.toml; export the token in your shell):
[mcp_servers.deployvera]
url = "https://mcp.deployvera.com/mcp"
bearer_token_env_var = "DEPLOYVERA_TOKEN"Windsurf and other clients: any client that supports remote MCP over Streamable HTTP works. Use the URL https://mcp.deployvera.com/mcp and, for tokens, the header Authorization: Bearer dvp_.... For stdio-only clients use npx -y deployvera-mcp with the DEPLOYVERA_TOKEN environment variable.
Tools
Tool | Scope | What it does |
| read | Projects with status, live URL, repo/branch, last build |
| read | Details, latest build, managed databases, and whether the live commit is behind the branch |
| read | Managed databases with engine, status and project. No credentials or connection strings |
| read | Tail (default 100, max 500) of a database's sanitized operational log |
| read | Build status (latest by default); |
| read | Tail (default 200, max 1000) of the sanitized build log, or runtime log |
| read | Reason code, plain-English cause, fix steps and the last log lines |
| env | Keys and flags only. Values are never returned |
| env | Bulk set |
| env | Delete a variable by key |
| deploy | Deploy the branch head, or a specific |
| deploy | Redeploy the retained image of an earlier build |
| deploy | Cancel a queued or running build |
| any | Status of an approval request. It cannot approve anything |
Example prompts
"Deploy my latest commit and tell me if it fails."
"Why did my last build fail? Fix it and redeploy."
"Which of my projects is behind its branch?"
"Add DATABASE_URL to the api project and restart it." (give the value only if you are comfortable sharing it with the agent)
"Show the last 100 runtime log lines of my site."
"Are my databases running? Show the Postgres log if one is not."
Approvals
Some actions need your approval in the browser, whatever the token or login scopes:
rollbackanddelete_env_varalways require it.Optionally, approve every deploy can be switched on for a token or login, so
deployrequires it too.
When approval is needed, the agent shows you a link that opens on https://deployvera.com/approve/…. Check that the address bar shows deployvera.com before you enter your password. Then check what will happen and confirm with your password and second factor. Requests expire after 10 minutes, and each approval works for one action only.
Security
Scopes:
read,deploy,env. A token only does what its scopes allow.Environment variable values are write-only: masked in
list_env_vars, never echoed byset_env_vars, scrubbed from error text. A value you type into a prompt is visible to that agent.Tokens and browser logins cannot create projects, change billing, domains or account settings, or reach admin routes. Use the panel for those.
Tokens expire after 30, 90 or 365 days and can be revoked at any time in Settings > AI agent access.
Rate limit: 120 requests per minute.
The hosted server is stateless and does not store tokens. Keep tokens out of repositories; prefer an environment variable or your client's secret prompt over a literal value in a config file.
Links
Website: https://deployvera.com
Documentation: https://deployvera.com/docs/en/quickstart
Privacy policy: https://deployvera.com/legal/privacy-policy
Terms of service: https://deployvera.com/legal/terms-of-service
Support: support@deployvera.com
License
MIT. See LICENSE. Logo: DeployVera mark, all rights reserved by DeployVera.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
Deploy apps from GitHub, read build/runtime logs, and manage your app's users from your AI agent.
Command center for developers coordinating AI coding agents across projects.
Project memory, documentation and controlled secret access for AI coding agents.
Deploy and manage your apps, databases, storage, and scheduled jobs from your AI agent
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceEnables AI agents to deploy code to any hosting provider by creating PRs, building, and verifying health checks, all from a single natural language command.55 npm1MIT

MileHost MCP Serverofficial
AlicenseCqualityCmaintenanceEnables AI coding agents to manage cloud containers, create and edit files, run commands, and deploy projects directly.247MIT- AlicenseNot gradedqualityBmaintenanceEnables AI assistants to build, deploy, and inspect apps through a hosted backend service, including reading platform docs, deploying raw source, inspecting logs, verifying live pages in browsers, and managing project tasks.MIT
- AlicenseNot gradedqualityCmaintenanceEnables checking deployability, planning deployments, viewing status and logs, and redeploying projects on your own cloud infrastructure directly from AI coding tools.85 npmMIT