Trust commands to be executed without manually clicking on RUN

## Problem

Currently, the Amazon Q extension for VS Code requires manual user confirmation (e.g., clicking "Run") for every suggested terminal command. This introduces workflow friction and degrades the speed advantage of using an AI assistant for terminal operations. For power users, repetitive tasks, or workflows within isolated and trusted local environments, this mandatory safety check becomes a bottleneck. There is no mechanism to opt out of this manual execution prompt within the IDE.

## Solution

Introduce a configurable setting in the Amazon Q VS Code extension to bypass the manual command execution requirement.

Configuration Toggle: Add a setting in the VS Code UI and settings.json (e.g., amazonQ.terminal.autoExecuteCommands) to enable or disable automatic command execution.

Immediate Execution: When enabled, commands generated and suggested by the Amazon Q Agent must automatically execute in the VS Code integrated terminal without requiring a secondary click or keystroke.

Granularity (Optional): Provide scope settings, allowing users to enable auto-execution globally, strictly within specific trusted workspaces, or exclusively for non-destructive, read-only commands (e.g., grep, ls, cat).

Start by locating the Amazon Q Agent terminal-command execution flow and the VS Code settings.json/UI registration mentioned in the issue. Trace where the manual "Run" confirmation is handled, then define the supported scope and safety behavior for the new toggle. Done means the setting is exposed and enabled commands execute automatically in the integrated terminal with the agreed scope and coverage.

Written by the indexing model from the issue text.