What would you like to be added?
Currently, Gemini CLI requires manual execution of prompts or commands. This limits its usefulness for workflows that require periodic monitoring, automation, or background AI tasks.
Other AI coding tools provide a loop-style command that allows the AI to run prompts automatically at defined intervals. This enables developers to automate tasks such as checking build status, monitoring logs, or reviewing pull requests without repeatedly issuing commands.
Gemini CLI lacks this capability, making it less efficient for long-running or monitoring-based workflows.
Why is this needed?
Introduce a /loop command that allows users to run prompts or commands repeatedly at a specified interval during an active CLI session.
Example syntax:
/loop 5m check if the deployment finished
Possible behavior:
Runs the prompt every specified interval (seconds, minutes, hours)
Continues until the session ends or the loop is manually stopped
Shows output after each iteration
Optional ability to stop with /loop stop
Optional enhancements:
Allow cron-style scheduling
Allow conditional stopping (e.g. stop when task succeeds)
Allow logging of loop results
Example Use Cases
CI/CD Monitoring
/loop 5m check the status of the latest GitHub Actions run
Log Monitoring
/loop 2m summarize new errors in server logs
Pull Request Review
/loop 10m check new PR comments and suggest fixes
Deployment Watcher
/loop 1m check if Kubernetes rollout completed
Expected Benefits
Enables lightweight automation workflows
Reduces repetitive manual prompts
Makes Gemini CLI more suitable for DevOps and continuous monitoring tasks
Improves parity with other AI coding assistants that support automated loops
Additional context
This feature would significantly improve Gemini CLI's usefulness as an AI-powered development assistant by enabling background task automation within the CLI session.
What would you like to be added?
Currently, Gemini CLI requires manual execution of prompts or commands. This limits its usefulness for workflows that require periodic monitoring, automation, or background AI tasks.
Other AI coding tools provide a loop-style command that allows the AI to run prompts automatically at defined intervals. This enables developers to automate tasks such as checking build status, monitoring logs, or reviewing pull requests without repeatedly issuing commands.
Gemini CLI lacks this capability, making it less efficient for long-running or monitoring-based workflows.
Why is this needed?
Introduce a /loop command that allows users to run prompts or commands repeatedly at a specified interval during an active CLI session.
Example syntax:
/loop 5m check if the deployment finished
Possible behavior:
Runs the prompt every specified interval (seconds, minutes, hours)
Continues until the session ends or the loop is manually stopped
Shows output after each iteration
Optional ability to stop with /loop stop
Optional enhancements:
Allow cron-style scheduling
Allow conditional stopping (e.g. stop when task succeeds)
Allow logging of loop results
Example Use Cases
CI/CD Monitoring
/loop 5m check the status of the latest GitHub Actions run
Log Monitoring
/loop 2m summarize new errors in server logs
Pull Request Review
/loop 10m check new PR comments and suggest fixes
Deployment Watcher
/loop 1m check if Kubernetes rollout completed
Expected Benefits
Enables lightweight automation workflows
Reduces repetitive manual prompts
Makes Gemini CLI more suitable for DevOps and continuous monitoring tasks
Improves parity with other AI coding assistants that support automated loops
Additional context
This feature would significantly improve Gemini CLI's usefulness as an AI-powered development assistant by enabling background task automation within the CLI session.