using-cline-cli

To safely run Cline CLI within a sandbox, you must start a guarded session that enforces sandbox policies and requires explicit approvals for escalated actions, network installs, or tool execution. This prevents unsafe operations while driving the model.

Switching providers and models from the Cline CLI involves following the guided selection workflow, which helps you choose appropriate inference providers and models for your specific task requirements while maintaining the sandbox guardrails.

The Cline CLI blocks commands and requires approvals to enforce security policies within the sandbox. It intercepts escalated actions, network installs, and specific tool usage rules for Execute, Read, Write, and Search to prevent unsafe or unauthorized actions.

Troubleshooting blocked commands and permission issues in Cline CLI involves following dedicated workflows that check sandbox approvals and policy constraints. It identifies which escalated action or network install triggered the block and guides you through the safe approval process.

Yes, you need to configure sandbox approvals before escalating commands in Cline CLI. The sandbox guardrails require explicit approval for escalated actions and network installs to ensure safe execution before you proceed with advanced tasks.

When using Cline CLI, specific tool usage rules apply to Execute, Read, Write, and Search operations to maintain sandbox security. These rules restrict unauthorized file modifications and command executions, requiring approvals for escalated actions.