checkBashCommand method
Check a bash command string for dangerous patterns.
Implementation
PermissionDecision checkBashCommand(String command) {
final (isDangerous, reason) = _checkBashDangers(command);
final request = PermissionRequest(
scope: PermissionScope.tool,
action: 'bash',
resource: command.length > 80
? '${command.substring(0, 80)}...'
: command,
toolName: 'Bash',
detail: isDangerous
? 'Bash command (WARNING: $reason): $command'
: 'Bash command: $command',
riskLevel: isDangerous ? RiskLevel.critical : RiskLevel.high,
metadata: {'fullCommand': command},
);
return _ruleSet.evaluateWithCache(request, _cache);
}