AI CodeOps for GitHub Copilot

GitHub Copilot works in your editor, as a cloud agent that opens pull requests, and as a reviewer on GitHub. AI CodeOps is a browser-based platform where AI agents plan, write, test and review a change, and the change becomes a pull request only after its quality gates pass and a person approves it. Connect Copilot to the AI CodeOps MCP server and Copilot can hand a task to a gated AI CodeOps session, read its findings and ask whether a pull request may ship.

What AI CodeOps gates, and what it doesn't

AI CodeOps gates the changes made in its own sessions. When the Copilot cloud agent works on a task, it pushes to its own branch and opens a pull request; GitHub's docs say it is subject to the repository's branch protections and required checks, and it can't approve or merge its own pull request (risks and mitigations). That pull request is not an AI CodeOps change: AI CodeOps doesn't review it, and the release check denies it because no AI CodeOps session made it.

To put the work through the AI CodeOps gates, have Copilot start an AI CodeOps session with the task, or start it in AI CodeOps yourself. The session writes the change, runs your tests, reviews it with several agents and opens the pull request on a codeops/run-<id> branch only when the gate passes (AI CodeOps for GitHub).

Connect Copilot in VS Code

1. In AI CodeOps, create a workspace API key under Configure → API Keys. It starts with sk- and is shown once. To let Copilot start sessions, choose Scope on the key and tick Start sessions from the MCP server (/mcp). See API keys.
2. Add the server to .vscode/mcp.json. VS Code's format uses a servers object, and an inputs entry prompts for the key so it is never written to the file (VS Code MCP configuration):

``json { "inputs": [ { "type": "promptString", "id": "aicodeops-key", "description": "AI CodeOps API key", "password": true } ], "servers": { "aicodeops": { "type": "http", "url": "https://console.aicodeops.dev/mcp", "headers": { "Authorization": "Bearer ${input:aicodeops-key}" } } } } ``

1. Start the server from VS Code, which asks for the key, then use Copilot in agent mode. It can call the six AI CodeOps tools.

Connect the Copilot cloud agent

The cloud agent reads an MCP configuration from the repository's settings (extend the cloud agent with MCP):

1. In the repository on GitHub, open Settings. Under Code, planning, and automation, choose Copilot, then MCP servers.
2. Add an Agents secret named COPILOT_MCP_AICODEOPS_KEY that holds the API key. Only secrets and variables whose names start with COPILOT_MCP_ are available to the configuration.
3. Paste this into MCP configuration and choose Save MCP configuration:

``json { "mcpServers": { "aicodeops": { "type": "http", "url": "https://console.aicodeops.dev/mcp", "headers": { "Authorization": "Bearer $COPILOT_MCP_AICODEOPS_KEY" }, "tools": ["list_projects", "get_session", "check_release_gate", "list_findings"] } } } ``

GitHub's docs say string fields in this configuration, headers included, can reference COPILOT_MCP_ secrets with $NAME or ${NAME}. The tools list is required; the example allows only the read tools, so the cloud agent can check the gate and read findings but can't start sessions. GitHub's docs also say Copilot code review can use MCP servers to pull context into a review (code review).

What the tools do

Each tool acts as the key's creator, with their role and budget, as in the console (MCP server guide).

Gate merges in CI

Your GitHub Actions workflow can ask the release gate directly and fail when it denies the pull request:

- name: Ask the AI CodeOps release gate
  env:
    AICODEOPS_API_KEY: ${{ secrets.AICODEOPS_API_KEY }}
    PR_URL: ${{ github.event.pull_request.html_url }}
  run: |
    curl -sf -X POST "https://console.aicodeops.dev/api/release/check" \
      -H "Authorization: Bearer $AICODEOPS_API_KEY" \
      -H "Content-Type: application/json" \
      -d "{\"pullRequestURL\":\"$PR_URL\"}" > gate.json
    jq -r '.summary' gate.json
    jq -e '.allowed == true' gate.json

Make the job a required status check in branch protection and a pull request merges only with an allowed verdict. That includes pull requests from the Copilot cloud agent, which the gate denies because no AI CodeOps session made them; use it on branches where AI CodeOps should be the path to merge. The fields, including a coverage floor and disallowed licenses, are on the API page.

Limits

• 120 MCP requests a minute from one address; 60 release checks a minute per workspace.
• The server is stateless and sends no notifications, so Copilot polls get_session to follow a session.
• Workspace API keys only. OAuth sign-in isn't supported yet.

Getting started

1. Create a free account: AI CodeOps is in open beta and runs in your browser.
2. Connect GitHub and create a project, using the quickstart.
3. Create an API key and add the server to VS Code or the cloud agent as above.

Comparing rather than combining? Read AI CodeOps vs GitHub Copilot code review and AI CodeOps vs GitHub Copilot coding agent.