jelly-droid-skill

Automate terminal-based coding tasks with multi-model CLI support.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/dev-jelly/jelly-dotclaude --skill jelly-droid-skill
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: jelly-droid-skill
Source: https://github.com/dev-jelly/jelly-dotclaude/tree/main/skills/jelly-droid-skill
Command: npx skills add https://github.com/dev-jelly/jelly-dotclaude --skill jelly-droid-skill

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Factory.ai Droid provides an autonomous AI coding agent that operates entirely in the terminal, enabling both interactive and non-interactive automation with multi-model support and session memory.

Core Features & Use Cases

  • Two Modes: Interactive full-screen UI and Exec non-interactive automation.
  • Autonomy Levels: 4 tiers from read-only to production.
  • Multi-Model: Supports Claude, GPT-5, and more.
  • Session Memory: Resume work exactly where you left off.
  • CI/CD Ready: Great for automated code reviews and pipelines.

Quick Start

Start an interactive session with droid, or run non-interactive edits with droid exec. For authentication, use Browser-based login or API keys as needed.

Frequently Asked Questions about jelly-droid-skill

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

FAQPage Schema
How do I automate coding tasks in the terminal with an AI agent?

Terminal-based autonomous AI coding automates code review, feature development, testing, and documentation. Droid provides interactive and non-interactive modes with multi-model support, session memory, and 4 autonomy tiers—from read-only to production—so you control how much the agent acts independently.

Can I use an AI coding agent in CI/CD pipelines?

Yes. Droid integrates with CI/CD workflows via non-interactive exec mode and connects to Jira, Notion, and Slack. Run automated code reviews and deployment workflows without manual intervention across multiple AI models.

What's the difference between interactive and non-interactive automation for coding tasks?

Interactive mode provides a full-screen terminal UI for real-time collaboration and control. Non-interactive exec mode runs automated edits silently in pipelines. Choose based on whether you need live oversight or background automation.

Do I need to configure multiple AI models, or does the agent pick one automatically?

Multi-model support lets you choose Claude, GPT-5, and others. The Skill includes model and context management controls so you can switch models and configure each session to fit your task requirements.

Can I resume an interrupted coding session where I left off?

Session memory preserves your work context, so you can resume exactly where the agent stopped. Interactive and non-interactive sessions both maintain state across interruptions.

What are the entry barriers to using a terminal-based AI coding agent?

Browser-based login or API key authentication gets you started quickly. No dependencies or prerequisites are required beyond terminal access and authentication credentials.