Connect your AI tool

Claude Code

1. Run the command below in your terminal.
2. Start Claude Code, type /mcp, choose teron and pick Authenticate.
3. Approve the connection on the Teron page that opens in your browser.

claude mcp add --transport http teron <your server URL>

Using an API key instead of a browser sign-in? Create one on Connections, set it as TERON_API_KEY and use: claude mcp add --transport http teron <your server URL> --header "Authorization: Bearer $TERON_API_KEY"

Claude Desktop

1. Open Customize, then Connectors.
2. Choose + Add, then Add custom connector.
3. Name it Teron and paste the server URL below, then choose Continue and Add.
4. Choose Connect and approve the connection on the Teron page that opens.

<your server URL>

Cursor

1. Use the install link, or add the JSON below to ~/.cursor/mcp.json (or .cursor/mcp.json in your project).
2. Open Cursor's MCP settings and choose Connect next to teron.
3. Approve the connection on the Teron page that opens in your browser.

{"mcpServers":{"teron":{"url":"<your server URL>"}}}

Codex CLI

1. Add the lines below to ~/.codex/config.toml.
2. Run codex mcp login teron in your terminal.
3. Approve the connection on the Teron page that opens in your browser.

[mcp_servers.teron]
url = "<your server URL>"

Using an API key instead of a browser sign-in? Create one on Connections, set it as TERON_API_KEY and use: [mcp_servers.teron] url = "<your server URL>" bearer_token_env_var = "TERON_API_KEY"

VS Code

1. Use the install link to add the server to VS Code.
2. Start the teron server from the MCP servers list and choose Allow when asked to sign in.
3. Approve the connection on the Teron page that opens in your browser.

{"name":"teron","type":"http","url":"<your server URL>"}

Using an API key instead of a browser sign-in? Create one on Connections, set it as TERON_API_KEY and use: {"inputs":[{"type":"promptString","id":"teron-key","description":"Teron API key","password":true}],"servers":{"teron":{"type":"http","url":"<your server URL>","headers":{"Authorization":"Bearer ${input:teron-key}"}}}}

Use with Roblox Studio

In Roblox Studio, open Assistant → … → Manage MCP Servers and turn on ‘Enable Studio as MCP server’, then connect the same AI tool to Studio using Studio’s quick connect. With both connected, ask your AI to start a Teron Engine project.

Studio’s MCP runs on your computer, so the AI that builds in Studio must run there too: a desktop or terminal AI app, not a chat in a browser tab.

Use with Unity (iOS and Android)

Your AI builds in the Unity editor, through a Unity MCP server or Unity's command line. Connect the same AI tool to it, then ask your AI to start a Teron Engine project for Unity mobile (or create one here and choose Unity mobile).

1. Install the Teron Engine Unity package once per Unity project. Your AI calls get_probe, which gives the install line: it adds com.teron.connect to Packages/manifest.json from a git URL ending in ?path=platform/unity/com.teron.connect. In Unity you can also add it with Window → Package Manager → + → Add package from git URL.
2. Claim a work order. It names the Unity project paths your AI may change (under Assets/Teron/, and Server/ for backend work) and the checks it must pass.
3. Send evidence. Your AI runs the probe from get_probe(kind="unity"): the menu Teron/Send Evidence in the open editor, the Unity MCP execute-code line, or Unity in batch mode. It checks compilation, runs the tests, measures every control on the iphone-15, iphone-se, ipad-11, android-phone and android-tablet profiles and captures each screen.
4. Upload the screenshots the probe lists, captioned "<scene> <profile>" (for example Match iphone-15), and server test results as a test_report.
5. Ask for a verdict. Teron Engine checks compile, tests, touch targets (44 pt on iOS, 48 dp on Android), the safe area and the layout on every device, the UI style and the concept art. You see each device's screenshot with its findings on the step.