Open Interpreter
Run Open Interpreter as a Subshell agent.
Before you start
Install interpreter using the vendor instructions and authenticate it under the execution account. Ask an administrator to enable the harness under Settings → Server Settings → Plugins in the server dashboard's sidebar. Follow Install an agent CLI to confirm detection on the target node.
Launch
- Open the new-subshell form using Launch a subshell, then choose Open Interpreter under Agent or terminal.
- Select a saved preset if you want a custom model. Configure that setting under Settings → Presets before launching.
- Choose the target node and an existing project directory.
- Launch and confirm that the CLI opens normally.
The plugin launches interpreter bare, which opens its interactive TUI. The preset model maps to the --model flag. It detects interpreter through PATH and its known locations, including ~/.openinterpreter/packages/standalone/current/bin/interpreter, where the standalone installer puts it. INTERPRETER_PATH can identify the binary explicitly in the execution environment. Before changing that override, see Files and paths for configuration locations and service-environment guidance.
MCP and session behavior
Register Subshell once under the OS account where Open Interpreter runs. In the server dashboard's sidebar, choose Settings → Presets → New preset → New agent preset, select Open Interpreter, and use the commands under Cross-subshell comms. Those commands use the MCP executable configured for your installation.
For a node with subshell on PATH, registration looks like this:
interpreter mcp add subshell -- subshell mcpOpen Interpreter takes the stdio command and its args after the -- separator, and it stores the registration under [mcp_servers] in its config home. Remove the registration later with interpreter mcp remove subshell.
If your preset sets INTERPRETER_HOME, Open Interpreter reads config.toml, including its MCP registration, from that directory instead of ~/.openinterpreter. Run the registration command with that same value on the execution machine so the registration lands where launch will read it, and use the same override when signing in or removing the registration. See Files and paths for environment configuration guidance.
The registration stores the command, not Subshell credentials. Open Interpreter's MCP child inherits them from the subshell where it runs, so one pane's identity stays inside that pane.
Presets support the model. Two flags change how much it asks before running a command: --ask-for-approval never runs commands without asking and returns execution failures to the model, and --approve-for-me routes approval requests through automatic review in the workspace-write sandbox. The -c override takes any config value as a dotted TOML path for a single run, for example -c model="gpt-5.1-codex". Continuing a previous conversation happens through a subcommand rather than a launch flag: interpreter resume opens a picker, and interpreter resume --last or interpreter resume <SESSION_ID> continues a session without the picker. The plugin doesn't provide automatic conversation resumption, attention hooks, or subagent tracking.
Subshell runs the real CLI rather than proxying its provider API. Provider configuration, usage limits, and CLI-specific permissions remain the agent's own settings.
Next steps
Read MCP configuration, Presets, and Agent and launch errors.
Edit on GitHubLast updated on
