TL;DR

Implementing a --dry-run flag for relevant Gemini CLI commands can provide a safe preview of command effects without making actual changes.

Guidance

• Identify commands that would benefit most from a dry run mode, such as file operations and automation tasks, to prioritize implementation.
• Design the output summary to clearly indicate that it's a simulation, including details of the actions that would be performed.
• Consider existing dry run or preview modes in other CLI tools (e.g., rsync, npm, git) for inspiration on implementation and user experience.
• Plan for potential extension to support custom extensions and MCP integrations, ensuring the dry run feature remains flexible and useful.

Example

# Example of how the --dry-run flag could be used
gemini cli-command --dry-run
# Output:
# "This is a dry run. The following actions would be performed:
# - Action 1
# - Action 2
# No changes will be made."

Notes

The implementation details of the --dry-run flag and its output will depend on the specific requirements and design of the Gemini CLI. It's essential to balance providing useful information with avoiding unnecessary complexity.

Recommendation

Apply a workaround by manually testing commands in a safe, non-production environment until the --dry-run feature is implemented, to mitigate the risk of accidental changes or data loss.