open-interpreter-runtime

name	open-interpreter-runtime
description	Open Interpreter runtime/session control: use Open Interpreter CLI % commands and shell flags to manage sessions and context in a worker terminal. Trigger when the controller needs to operate Open Interpreter like a human.

Open Interpreter Runtime

Overview

Operate the Open Interpreter CLI safely: manage session state, verbosity, and context from a worker terminal.

Session Safety

1. Confirm idle state

• Snapshot and/or status the worker; do not intervene mid-run.
• Only proceed when the worker is at a prompt or explicitly idle.

2. Prefer in-session resets

• Use %reset to reset the conversation state.
• Use %undo to revert the last step.
• Use %retry to rerun the last step.

Core % Commands (Open Interpreter)

Use %help to see the full list. Common commands:

• %reset reset the conversation.
• %undo undo the last action.
• %retry retry the last action.
• %verbose toggle verbose output.
• %tokens show token usage.
• %editor open the editor for the current session.
• %history / %messages view conversation history.
• %logs / %debug show logs and debug info.
• %multi_line / %paste handle multi-line input.
• %profiles / %profile manage profiles.
• %list list available commands.
• %clear clear the screen.

Launch (Shell)

• interpreter starts the CLI.
• interpreter --help shows available flags and options.

Guardrails

• Do not restart mid-run.
• Prefer %reset or %undo over full restarts when possible.
• If the worker is not Open Interpreter, switch to the model-specific runtime skill instead.