This repository provides a curated starter kit for modern, agentic software development, offering practical agent skills, configuration templates, and security checklists. It targets development teams seeking a controlled and efficient entry into AI-assisted workflows, enabling faster planning, development, testing, reviewing, and securing of code using tools like Claude Code, Codex, Cursor, and Gemini CLI.

How It Works

The project aggregates field-tested agent skills, configuration templates, and security guidelines distilled from agentic-coding.ch. It organizes these resources into distinct directories (skills/, templates/, SECURITY.md) to facilitate a structured approach. The core advantage lies in its emphasis on a "cleanly controlled entry," prioritizing security reviews and isolated testing before productive use of AI-driven development tools.

Quick Start & Requirements

To begin, select a skill from the provided table or the skills/ directory. It is crucial to verify the source, license, and maintainer of the upstream repository for each skill. Subsequently, review the security checklist in SECURITY.md and install the skill in a test project or isolated environment before productive deployment. Ensure behavior, permissions, and network access are plausible. Prerequisites include compatibility with AI coding tools such as Claude Code, Codex, Cursor, and Gemini CLI, and likely Node.js for npx commands. Links to specific skills and templates are available within the repository.

Highlighted Details

• Diverse Agent Skills: Features curated skills for tasks including persistent planning, web scraping, structured development workflows, UI/UX/accessibility guidelines, security audits (CodeQL, Semgrep), SaaS integrations, programmatic video generation, codebase search, CI error correction, and Gemini API development.
• Configuration Templates: Includes templates for CLAUDE.md (Second Brain, Security Testing, Code Review, Full-Stack Development), ChatGPT/OpenAI, Google Gemini, and Gemma 4 local configurations.
• Security-First Approach: Offers a dedicated SECURITY.md checklist to vet agent skills, emphasizing safe handling of shell commands, network access, and sensitive data.
• Multi-AI Tool Support: Designed for interoperability with various AI coding assistants, including Claude Code, Codex, Cursor, and Gemini CLI.

Maintenance & Community

This repository is copyrighted by Dr. René Bäder (PhDs). No specific details regarding active maintenance, community channels (like Discord/Slack), or a public roadmap are provided in the README.

Licensing & Compatibility

The starter kit is licensed under the GNU General Public License v3.0 or later (GNU GPL v3.0+). This strong copyleft license may impose obligations on derivative works, potentially restricting integration into closed-source or proprietary projects without careful legal review. Individual third-party skills included may carry their own, separate licenses.

Limitations & Caveats

Agent skills included in this kit possess the capability to execute shell commands, browser automation, and network operations, posing potential security risks. Thorough vetting via the SECURITY.md checklist is mandatory before installation, particularly in production or regulated environments. The project relies on external, third-party skills, each with its own licensing, maintenance status, and security considerations. This repository serves as a curated collection and guide, requiring users to manage and integrate individual skills.