waybarrios/opencode-power-pack

54 rigorous skills for Codex, OpenCode, and Pi: code review, security audit, feature development, frontend design, MCP tools, Hugging Face ML/training, and more.

What it solves

OpenCode Power Pack provides a portable collection of 54 specialized workflows (skills) for AI coding agents like Claude Code, Codex, OpenCode, and Pi. It solves the problem of fragmented AI agent workflows, allowing developers to use consistent, high-quality prompts and orchestration patterns for tasks like security audits, feature development, and ML model training across different agent platforms.

How it works

The project distributes these workflows as plugins or packages. It includes a CLI tool for selective installation into the shared .agents/skills directory used by various agents. To enhance security, it features a native sandbox runner that uses OS-level isolation (Seatbelt on macOS and Bubblewrap on Linux) to execute agent-triggered commands with least-privilege profiles (e.g., observe, develop, network-read, publish).

Who it’s for

Software engineers, security researchers, and ML practitioners who use AI coding agents and want a standardized set of professional-grade workflows for code review, security analysis, and Hugging Face ecosystem integration.

Highlights

  • Multi-Agent Support: Compatible with Claude Code, Codex, OpenCode, and Pi.
  • Extensive Skill Catalog: 54 workflows covering code review, security threat modeling, frontend design, and technical paper summarization.
  • Native Sandboxing: Opt-in, fail-closed execution environment to prevent unauthorized host access during agent command execution.
  • Hugging Face Integration: Dedicated skills for model discovery, local inference, fine-tuning (SFT/DPO/GRPO), and Spaces deployment.
  • Least-Privilege Profiles: Granular control over workspace access, network connectivity, and external side effects.

Related

  • Project
    finbarr/yolobox

    A containerized sandbox for AI coding agents that allows them to run commands with sudo privileges while protecting the host's home directory from accidental destruction.

  • Project
    opensandbox-group/OpenSandbox

    A general-purpose sandbox platform for AI applications that provides secure, isolated environments for code execution, agent evaluation, and RL training.

  • Project
    nikvdp/cco

    A sandboxing wrapper for Claude Code and other AI coding agents that provides secure isolation using native OS tools or Docker without sacrificing autonomous productivity.

  • Project
    jamditis/claude-skills-journalism

    A genuine open‑source repo that bundles reusable “Agent Skills” for Claude Code (and Codex) to help journalists and newsroom teams automate tasks such as fact‑checking, source verification, FOIA filing, data‑journalism, PDF design, and multi‑agent workflows. Install via Claude’s plugin commands or copy individual `SKILL.md` folders; the repo includes a catalog, docs site, and up‑to‑date skill set.

  • Project
    mxyhi/ok-skills

    A curated collection of reusable skills and playbooks for AI coding agents, providing portable instructions for tasks like browser automation, documentation lookup, and codebase planning.