steipete/agent-scripts

Scripts for agents, shared between my repositories.

What it solves

It provides a centralized management system for AI agent instructions, reusable skills, and helper scripts across multiple local workspaces. It eliminates the need to manually duplicate rules and capabilities for different AI agents (like Codex and Claude Code) across various projects.

How it works

The project uses a canonical repository to store shared rules (AGENTS.MD) and skills (defined in SKILL.md files with YAML front matter). A synchronization script (sync-skills) creates symlinks in the agent-specific configuration directories (e.g., ~/.codex/ and ~/.claude/) to map these shared resources to the agents. It supports both whole-root links for Codex and flat per-skill links for Claude Code, while handling name collisions and pruning stale links.

Who it’s for

Developers who use multiple AI coding agents across various local projects and want a single source of truth for their agent's behavior, custom skills, and utility scripts.

Highlights

  • Cross-Agent Syncing: Automatically configures instructions and skills for both Codex and Claude Code.
  • Skill Routing: Uses YAML front matter in skill files to provide short trigger phrases for agent routing.
  • Dependency-Light Helpers: Includes portable scripts and a standalone Chrome DevTools helper for browser automation.
  • Validation: Includes a validate-skills script to ensure skill definitions are correctly formatted.

Related

  • Project
    openai/skills

    A repository of reusable instruction and script packages that AI agents can use to perform specific tasks in a repeatable way.

  • Project
    openclaw/agent-skills

    OpenClaw Agent Skills is a reusable library of workflow “skills” (e.g., autoreview, behavior validation, session viewer) for AI coding agents in OpenClaw projects. It provides a Python installer to symlink or copy these skills into a user’s agent‑skill directory, enabling consistent, version‑controlled automation across repositories. Includes validation scripts, MIT licensing, and support for Codex/Claude agents.

  • Project
    MengTo/Skills

    A curated library of reusable, procedural 'skills' and workflows designed to help AI coding agents build consistent web interfaces, 3D effects, and game systems.

  • Project
    Kotlin/kotlin-agent-skills

    A collection of standardized AI agent skills that provide specialized instructions and resources to help AI coding agents perform Kotlin-specific development tasks.

  • Project
    dongshuyan/compass-skills

    COMPASS Skills is an alpha‑stage, modular toolkit of ten markdown‑based AI agent skills (task clarification, task graph, pause‑resume, hand‑off prompt, user profile, skill‑builder/upgrader, clear communication, academic humanizer, interview assessment) that let LLM‑driven agents persist context across conversations, align with user preferences, and safely evolve from real run evidence. Install via `npx skills add …` or copy the skill folders into a harness’s skill directory.