How does the sandboxed execution pattern work?

It provides standardized patterns for using Gemini CLI flags to execute potentially destructive or untrusted code within isolated environments like Docker or Podman.

When should I delegate a task from Claude to Gemini?

You should delegate to Gemini when you need to handle interactive TUIs, run risky commands in a sandbox, analyze massive codebases (1M+ tokens), or require native GCP/Firebase integration.

Can this skill help with interactive commands like Vim or Git Rebase?

Yes, it includes an Interactive Handoff pattern specifically for TUI commands where Claude identifies the requirement and initiates a Gemini session with PTY support.

Which Gemini model is recommended for bulk analysis?

Gemini 2.5 Flash is recommended for bulk analysis and processing large files due to its high context capacity and lower cost, whereas Pro is reserved for complex reasoning.

Gemini Delegation Patterns

Name: Gemini Delegation Patterns
Author: melodic-software

bymelodic-software

•

生産性とワークフロー

Optimizes cross-agent workflows by determining when and how to delegate complex tasks from Claude to the Gemini CLI.

概要

This skill provides a strategic framework for managing delegation between Claude and Gemini, allowing developers to leverage the unique strengths of each agent within the Claude Code environment. It features a comprehensive decision matrix for scenarios such as interactive TUI requirements, high-risk command execution, and large-scale codebase analysis exceeding 1M tokens. By standardizing execution patterns—including sandboxed testing, checkpointed experimentation, and bulk analysis—this skill ensures efficient task distribution, native Google Cloud integration, and robust error handling for sophisticated multi-agent development workflows.

主な機能

Context-aware decision matrix for Claude vs. Gemini selection
Interactive TUI handoff patterns for tools like Vim and Git Rebase
Sandboxed execution patterns for high-risk or untrusted commands
Model selection guidance for Gemini Flash and Pro optimization
1 GitHub stars
Strategies for analyzing massive codebases up to 1M+ tokens

ユースケース

Executing risky shell commands and migrations within isolated sandboxes
Analyzing monolithic codebases that exceed standard AI context limits
Performing interactive terminal tasks like Git rebasing via Gemini PTY

概要

主な機能

Context-aware decision matrix for Claude vs. Gemini selection
Interactive TUI handoff patterns for tools like Vim and Git Rebase
Sandboxed execution patterns for high-risk or untrusted commands
Model selection guidance for Gemini Flash and Pro optimization
1 GitHub stars
Strategies for analyzing massive codebases up to 1M+ tokens

ユースケース

Executing risky shell commands and migrations within isolated sandboxes
Analyzing monolithic codebases that exceed standard AI context limits
Performing interactive terminal tasks like Git rebasing via Gemini PTY