Back to directory

Claude Code Skills & Agents

A curated marketplace plugin of skills, sub-agents and configuration templates for Claude Code, adding deep research, image generation and GitHub automation with multi-model switching.

Claude Code Skills & Agents

Claude Code Skills & Agents is a curated collection of skills, sub-agents and configuration templates that extend Claude Code with capabilities such as deep research, image generation, and GitHub issue and pull-request automation, alongside templates for switching between several model providers. It installs as a single marketplace plugin from inside a Claude Code session or from its command line, or as individual skills through the skills command-line installer.

The bundled skills include brainstorming, a structured, question-by-question requirements gathering process before implementation begins, a Codex-configuration skill, deep research, GitHub issue-fixing and pull-request review, image generation, a skill-authoring helper, translation, and YouTube transcript summarizing. Manual installation clones the repository over an existing Claude Code configuration folder and can route requests through a proxy such as a Copilot gateway or a LiteLLM proxy server. It is written in Python and licensed under MIT.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.