This project automates the creation of intelligent agents from natural language descriptions, transforming repetitive manual workflows into self-improving automated processes. Targeting business owners, professionals, and researchers, it promises significant ROI by drastically reducing task times and increasing productivity.

How It Works

The Agent Creator acts as a meta-skill within Claude Code, guiding the generation of "Claude Skills" via a rigorous 6-phase process informed by a comprehensive /references knowledge base. It supports simple and complex agent architectures, using a "-cskill" suffix. Version 2.1+ integrates AgentDB for progressive learning, while an enhanced 4-layer activation system ensures high reliability (99.5%+) for created agents, supporting multi-intent queries.

Quick Start & Requirements

Installation: /plugin marketplace add FrancyJGLisboa/agent-skill-creator in Claude Code. Prerequisites: Claude Code CLI, Python 3.8+ for generated agents. Optional: AgentDB CLI and Python libraries. Agent creation: 15-90 minutes.

Highlighted Details

• Automated Agent Generation: Creates production-ready agents from plain language, eliminating manual coding.
• High-Reliability Activation: 4-layer system (Keywords, Patterns, NLU, Context Filtering) achieves 99.5%+ reliability and handles multi-intent queries.
• Progressive Learning (v2.1+): AgentDB integration enables agents to learn from usage, improving creation speed and offering personalized suggestions.
• Cross-Platform Export (v3.2): Skills exportable for Claude Code, Desktop, Web, and API platforms.
• Production-Quality Output: Agents include substantial Python code, extensive documentation, error handling, testing, and one-command installation.

Maintenance & Community

Actively developed with a roadmap to Q2 2026, including AI-powered template generation and an agent marketplace. Community support via GitHub Discussions/Issues.

Licensing & Compatibility

Apache 2.0 license: free to use, modify, and distribute, including commercially.

Limitations & Caveats

Advanced customization may require technical expertise. Full learning features depend on optional AgentDB integration. Agent creation time varies with complexity. Project is under rapid development.