Deploy from VS Code: Copilot agent sessions and the Livops MCP server
In VS Code, add Livops to the workspace's .mcp.json as an http server at https://livops.uz/api/mcp, select the Copilot harness in a chat session, and complete the sign-in when prompted. Copilot can then deploy, read build logs, explain a failure and roll back, and each tool call still needs your approval.
By Livops team, the people who build and run LivopsUpdated
GitHub Copilot in VS Code can call MCP tools from a chat session. With Livops connected, the same session that changes your code can deploy it, follow the build and roll it back. This page has the config for VS Code’s current Copilot format, the older one it replaces, both ways of signing in, and how to keep Copilot to the tools you choose.
What you need
- A Livops account, and a project or a GitHub repository to make one from.
- VS Code with Copilot, and the Copilot harness selected in a chat session[1].
- For the token route only: an
lvp_API token from the team’s Settings → API tokens, created by the team owner.
Livops is a Streamable HTTP MCP server at https://livops.uz/api/mcp.
Which config file?
VS Code now has two formats. The Copilot harness reads a top-level mcpServers object from a workspace-root .mcp.json, or from ~/.copilot/mcp-config.json for every workspace[1]. The editor-managed servers format in .vscode/mcp.json belongs to the Local harness, which VS Code’s reference calls deprecated and due to be removed in February 2027; for new Copilot configurations it says to use the native files instead[1]. Both are below.
You can also run MCP: Add Server from the Command Palette for a guided flow[2].
Connect through the browser
.mcp.json in the workspace root:
{
"mcpServers": {
"livops": {
"type": "http",
"url": "https://livops.uz/api/mcp",
"tools": ["*"]
}
}
}
type: "http" is Streamable HTTP, and ["*"] makes all of the server’s tools available[1]. VS Code’s reference says to complete authorization when prompted and to use Authenticate in the session’s server actions when it is offered[1]. Livops then opens in the browser: the app’s name, a team to pick, what the app will be able to do, and Allow. With two-factor on, your code comes first.
The app gets a key that lasts an hour, works only at /api/mcp and renews itself. It shows under Account → Connected AI apps, where Disconnect ends it.
Connect with an API token
{
"mcpServers": {
"livops": {
"type": "http",
"url": "https://livops.uz/api/mcp",
"headers": { "Authorization": "Bearer ${LIVOPS_TOKEN}" },
"tools": ["*"]
}
}
}
In this format, header values can reference environment variables as ${NAME}; the braced form is required, and the value comes from the environment of the host that connects to the server[1]. Input prompts are not supported here[1], so set LIVOPS_TOKEN before VS Code starts. The reference also warns against literal credentials in shared files[1], and .mcp.json sits in the workspace you commit.
The older .vscode/mcp.json
If you still use the Local harness, its format is servers, and it can ask for the token once and hide what you type[1]:
{
"inputs": [
{
"type": "promptString",
"id": "livops-token",
"description": "Livops API token",
"password": true
}
],
"servers": {
"livops": {
"type": "http",
"url": "https://livops.uz/api/mcp",
"headers": { "Authorization": "Bearer ${input:livops-token}" }
}
}
}
What can Copilot do in Livops?
It works in one team, as you, with your role. A viewer only reads.
| Browser sign-in | lvp_ API token |
|
|---|---|---|
| Read projects, deploys, build and runtime logs, metrics | Yes | Yes |
| Deploy, roll back, restart, set variables | Yes | Yes |
| Delete anything | Yes, if your role allows | No |
| Read a database password or bucket key | Yes, if your role allows | No |
| Open a Payme or Click top-up link, redeem a promo code | Yes | No |
Card numbers, your password, two-factor, sessions, creating API tokens, deleting the account, handing a project over, project members and raw backup downloads have no tool at all.
Expose fewer tools, approve the rest
The tools field takes names, and listing them limits what the server offers; it does not bypass tool approvals[1]. A Copilot that should watch builds but never change anything can have:
"tools": ["list_projects", "list_deploys", "get_deploy_log", "get_logs", "get_metrics"]
For the tools you do expose, review approval requests as they appear and approve only actions you understand[2]. Livops marks each tool with readOnlyHint or destructiveHint; the specification asks clients to keep a human able to deny a call, and to treat annotations as untrusted unless the server is trusted[3].
Build logs and commit messages reach the model as data. A line in a log that gives the model instructions is still part of a log; your approval is what decides whether anything happens.
Deploying from VS Code on Livops
- Add the
livopsentry to.mcp.jsonin the workspace root, with just the URL. - Open a chat session with the Copilot harness, complete the sign-in, pick the team, press Allow.
- Ask: “List my Livops projects.”
- Push to the branch the project follows (every push to it deploys), and ask Copilot to follow the build with
get_deploy_log. If the build fails, the previous version keeps running, andexplain_failureasks Livops’s assistant why.
Every tool is listed in MCP (AI clients); what a token may do is in API tokens.
Questions
Which file does VS Code read for MCP servers?
For Copilot sessions, VS Code's reference points new configurations at a workspace-root .mcp.json with a top-level mcpServers object, or ~/.copilot/mcp-config.json for every workspace. The older .vscode/mcp.json with a servers object belongs to the Local harness, which VS Code says is deprecated.
Do I need a token for VS Code?
No. Add just the address and complete the authorization when VS Code prompts you; Livops opens in the browser, you pick a team and press Allow. A token is for machines without a browser.
Can I give Copilot only some of the Livops tools?
Yes. The tools field takes tool names. ["*"] exposes all of them; a list such as list_projects, get_deploy_log and get_logs exposes only those. It limits what Copilot sees and does not bypass tool approvals.
How do I keep the token out of .mcp.json?
Write the header as Bearer ${LIVOPS_TOKEN} and set LIVOPS_TOKEN in the environment VS Code runs in. The braced form is required; a bare $LIVOPS_TOKEN is read as literal text.
What can a token not do?
Delete anything, read a database password or a bucket key, pay, or change the variables a whole environment shares. Livops refuses those for an lvp_ token on the server.
How do I disconnect VS Code?
In the cabinet, Account, then Connected AI apps, then Disconnect. With a token, revoke it under Settings, API tokens.
Sources
- VS Code — MCP configuration reference (code.visualstudio.com). Checked .
- VS Code — Add and use MCP servers in VS Code (code.visualstudio.com). Checked .
- Model Context Protocol — Tools (specification 2025-11-25) (modelcontextprotocol.io). Checked .
Put your project online today
Connect GitHub or drop a folder. No card is asked for.
Connect VS CodeRelated
- MCP server for deploying apps: what to look forAn MCP deploy server lets Claude Code, Cursor, Codex or Copilot deploy, read logs and roll back on your hosting account. Judge one by how it signs in, how far a key reaches, whether its tools say which ones change things, and what you can see and revoke afterwards. Vercel, Render, Railway, Netlify and Livops all ship one.
- Deploy from Cursor: connect Livops over MCPAdd Livops to ~/.cursor/mcp.json with just its address, https://livops.uz/api/mcp, and sign in through the browser. Cursor's agent can then deploy, read build logs, explain a failure and roll back, and it asks you before each tool call unless you change that.
- Deploy from Claude Code: connect Livops over MCPRun one command, claude mcp add --transport http livops https://livops.uz/api/mcp, then sign in through /mcp. From then on Claude Code can list your projects, deploy, read the build log, explain a failure and roll back, as you and only in the team you picked.
- What Livops's AI assistant does when a build failsWhen a build fails on Livops, the previous version keeps running and one button asks the assistant why. It reads the failed build's log and your project's commands, says what went wrong and how to fix it, and may offer one fix as a button that changes nothing until you press it. It never deletes, never sees variable values, and does not edit your code.