Overview
This is a comprehensive tracking issue for making Gemini CLI's non-interactive mode production-ready and suitable for automation, CI/CD pipelines, and programmatic integration. While Gemini CLI excels in interactive use, its non-interactive mode currently has several limitations that prevent reliable automation and integration with other tools.
Problem Statement
Non-interactive mode is essential for:
- Automations: Batch processing and scripted workflows e.g. code reviews
- Third-party Integrations: Building tools and services on top of Gemini CLI
However, the current non-interactive mode suffers from:
- Critical Issues: Unreliable Error Handling (#6745), Tool Execution Failures (#6770), Input Processing Bugs (#8015)
- Integration Limitations: Limited Output Formats (#8022), Missing Command Support (#5435)
- Operational Pain Points: MCP Server Management (#4372), Session Continuity (#1530), File Input Limitations** (#3311)
Goals
- Feature-Complete: Support for all relevant features available in interactive mode
- Reliable & Predictable: Consistent behavior with proper error handling and exit codes
Impact
Completing this work will:
Immediate Benefits
- Unlock GitHub Actions Integration: Enable reliable CI/CD workflows with proper error handling
- Reduce Support Burden: Fewer automation failures and clearer debugging information
- Enable Enterprise Adoption: Production-ready automation capabilities for large organizations
- Improve Developer Experience: Consistent, predictable behavior across interactive and non-interactive modes
Strategic Benefits
- Ecosystem Growth: Enable third-party tools and integrations to build on Gemini CLI
- Foundation for SDK: Patterns and APIs developed here will inform future SDK development
- Usage: Open new use cases in automation, DevOps, and enterprise workflows
Relationship to Future SDK Plans
This work is complementary to, not competing with, planned SDK development:
Bridge Strategy
- Immediate Solution: Solves urgent automation needs while SDK is in exploration
- Foundation Building: JSON output schema and error handling patterns become SDK building blocks
- User Validation: Real-world usage informs SDK design decisions
- Migration Path: Users can gradually move from CLI-based to SDK-based solutions when ready
Different Use Cases
- CLI Non-Interactive: Shell scripting, CI/CD pipelines, GitHub Actions, command-line automation
- Future SDK: Direct programmatic integration, embedded applications, complex workflows with fine-grained control
Note: This tracking issue will be updated as work progresses and new requirements emerge. Progress is tracked through the linked sub-issues, with regular updates posted here.
Overview
This is a comprehensive tracking issue for making Gemini CLI's non-interactive mode production-ready and suitable for automation, CI/CD pipelines, and programmatic integration. While Gemini CLI excels in interactive use, its non-interactive mode currently has several limitations that prevent reliable automation and integration with other tools.
Problem Statement
Non-interactive mode is essential for:
However, the current non-interactive mode suffers from:
Goals
Impact
Completing this work will:
Immediate Benefits
Strategic Benefits
Relationship to Future SDK Plans
This work is complementary to, not competing with, planned SDK development:
Bridge Strategy
Different Use Cases
Note: This tracking issue will be updated as work progresses and new requirements emerge. Progress is tracked through the linked sub-issues, with regular updates posted here.