gemini-reference

Start, resume, and review Gemini CLI sessions with JSON outputs.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/sirn/dotfiles --skill gemini-reference
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gemini-reference
Source: https://github.com/sirn/dotfiles/tree/main/var/agents/skills/gemini-reference
Command: npx skills add https://github.com/sirn/dotfiles --skill gemini-reference

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides a reference for calling the Gemini CLI agent from other agents, enabling structured task delegation and session-aware interactions.

Core Features & Use Cases

  • Central guidance on starting, resuming, and reviewing Gemini sessions.
  • Supports deterministic, JSON-formatted outputs for reliable downstream processing.
  • Useful for coordinating multi-step tasks across agents that rely on Gemini for reasoning.

Quick Start

Start a new Gemini task by creating a session file and issuing a read command to generate a structured JSON response.

Frequently Asked Questions about gemini-reference

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I call the Gemini CLI from another agent for task delegation?

To call the Gemini CLI from another agent, you use a reference structure that standardizes session-based communication, creating a session file and issuing a read command to generate structured JSON responses for reliable downstream processing.

How does session management work when automating Gemini CLI tasks?

Session management for Gemini CLI tasks works by using a file-based session store to start, resume, and review multi-turn prompts, ensuring deterministic results across complex reasoning workflows.

Can I get structured JSON output from a Gemini CLI session for downstream processing?

Yes, you can get structured JSON output from a Gemini CLI session. The automation enforces a stable JSON output format to ensure deterministic results for reliable downstream agent processing.

Do I need a specific environment setup to resume multi-turn prompts with the Gemini CLI?

Yes, resuming multi-turn prompts with the Gemini CLI requires a working file-based session store and Gemini tooling to maintain session state and ensure deterministic results across interactions.

What is the best way to coordinate multi-step tasks across agents using Gemini?

The best way to coordinate multi-step tasks across agents using Gemini is to standardize session-based communication, allowing agents to start, resume, and review tasks through deterministic JSON-formatted outputs.