Overview
Sandboxing controls the agent’s ability to interact with your system. It determines which tools are available, which operations need approval, and what actions are blocked entirely.Sandbox Modes
- Unrestricted
- Sandboxed (Default)
- Locked
- All tools available
- No approval required
- Suitable for: trusted personal environments, development
Tool Policies
Fine-grained policies override the global sandbox mode for individual tools:Blocked Patterns
TheblockedPatterns array uses substring matching against the command string. If any pattern matches, the command is rejected without asking for approval.
Common patterns to block:
| Pattern | Reason | |
|---|---|---|
rm -rf / | Prevents recursive deletion of root | |
sudo | Prevents privilege escalation | |
chmod 777 | Prevents world-writable permissions | |
mkfs | Prevents filesystem formatting | |
> /dev/ | Prevents writing to device files | |
| “ :(){ : | :& };: “ | Fork bomb |
Approval Flow
When an operation requires approval:- In the Dashboard
- In Channels
A modal appears with the command or action details. The user clicks “Approve” or “Reject.”
If no approval is received within a configurable timeout (default: 5 minutes), the operation is rejected automatically.
Path Restrictions
TheallowedPaths array restricts where the exec tool can operate:
