Skip to content

Repository files navigation

atlas

A growing, open collection of agentic workflows, skills, and agents — the building blocks of a personal operating system for working with AI agents.

Everything here is portable Agent Skills (the open SKILL.md standard originated by Anthropic and adopted across the ecosystem). Skills are plain markdown with a little frontmatter, so they drop into Hermes Agent, Claude Code (~/.claude/skills/), and any other agent that speaks the format.

Why "atlas"

An atlas is a collection of maps — each one self-contained, all of them part of one navigable whole. That's the intent here: a personal system that grows one useful capability at a time, where each piece stands on its own but they compound into something larger.

What's inside

atlas/
├── skills/        # SKILL.md capabilities — reusable procedures an agent loads on demand
├── agents/        # agent definitions / personas (coming as the collection grows)
└── workflows/     # multi-step orchestrations that chain skills + agents (coming)

Skills

Skill What it does
pixar-story-spine Turn a complex work problem, design, or decision into a Pixar-style narrative using Kenn Adams' Story Spine — with anti-slop guardrails so the audience stays the hero and the "so what" is measurable.

More landing over time. Agents and workflows will populate their directories as the system fills in.

Install a skill

Each skill is a self-contained directory. To use one, copy it into your agent's skills folder:

Hermes Agent

git clone https://github.com/dstkwll/atlas.git
cp -r atlas/skills/pixar-story-spine ~/.hermes/skills/creative/

Claude Code

cp -r atlas/skills/pixar-story-spine ~/.claude/skills/

Your agent auto-discovers the skill on its next session and loads it when a prompt matches the skill's trigger description.

Structure & conventions

Every skill follows the same shape so the collection stays consistent and portable — see CONTRIBUTING.md for the full convention. In short:

  • One directory per skill under skills/<skill-name>/, containing a SKILL.md.
  • YAML frontmatter with name, description (with trigger phrases), version, author, license, platforms.
  • Body: when to use / when NOT to use, a numbered procedure, an anti-pattern or pitfalls section, and a worked example.
  • Attribution preserved for anything derived from upstream work.

License

MIT. Individual skills preserve upstream attribution in their own SKILL.md where they build on prior work.

About

A growing collection of agentic workflows, skills, and agents — the building blocks of a personal OS for AI agents.

Resources

Contributing

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages