Skills (Agentic Simulation + Tailscale Utilities)

Repository: github.com/sgoley/skills

skills is a broader evolution of the earlier Tailscale-only skill pack: one repository that combines high-stakes decision and simulation workflows with practical infrastructure utility skills.

What it does

installs as a single package via gh skill install sgoley/skills
provides reusable skill specs for both strategic preparation and day-to-day operations
organizes guidance into two categories: agentic-game-simulation/ and utils/
includes structured workflows, required output formats, and practical command examples
is designed for agents that support the Agent Skills format

Repository structure

skills/
├── agentic-game-simulation/
│   ├── write-character-sheet/
│   │   ├── SKILL.md
│   │   ├── REFERENCE.md
│   │   └── EXAMPLES.md
│   └── game-theory-scenario/
│       ├── SKILL.md
│       ├── REFERENCE.md
│       └── EXAMPLES.md
└── utils/
    ├── tailscale-cli/SKILL.md
    ├── tailscale-dns/SKILL.md
    ├── tailscale-exit-nodes/SKILL.md
    ├── tailscale-funnel/SKILL.md
    ├── tailscale-serve/SKILL.md
    └── tailscale-ssh/SKILL.md

Core category: agentic-game-simulation

`write-character-sheet`

A structured self-profiling workflow that runs an interview, captures constraints and context, and converts responses into evidence-weighted scores plus narrative guidance. Output includes core and derived stats, habits to keep/change, goals by horizon, resources/support, influential inputs, and daily structure.

`game-theory-scenario`

A scenario-planning workflow for high-stakes interactions. It interviews for context and artifacts, maps actor incentives and pressure points, builds 3-6 branch pathways (cooperative, neutral, adversarial, mixed), and outputs a readiness pack with response menus, risk/escalation plans, and rehearsal drills.

Utility category: Tailscale operations

Skill	Focus
`tailscale-cli`	Device lifecycle, diagnostics, authentication, routes, certs, and core CLI workflows
`tailscale-dns`	MagicDNS, nameservers, split DNS, search domains, and DNS troubleshooting
`tailscale-exit-nodes`	Advertising and using exit nodes, LAN access options, and security tradeoffs
`tailscale-funnel`	Public sharing via encrypted tunnels, policy prerequisites, cert requirements, and port constraints
`tailscale-serve`	Internal tailnet sharing of files/services with status and access patterns
`tailscale-ssh`	SSH enablement, ACL-based authorization, check mode, and operational troubleshooting

End-to-end workflow

Build the player model with /write-character-sheet.
Build the scenario map with /game-theory-scenario.
Stress-test likely branches and prepare fallback moves.
Execute in the real world, then iterate your model.
Use utility skills (especially Tailscale) for secure networking and service exposure when needed.

Why this project matters

Many assistants are good at abstract strategy or concrete commands, but not both in one reusable skillset. This repository intentionally combines:

human systems modeling (incentives, behavior, constraints, branch logic)
infrastructure execution (networking, access control, DNS, service sharing)

That makes skills useful as both a planning framework and an operator runbook library, with the Tailscale set now one utility slice of a larger agentic toolkit rather than the whole project.