Agents

40 Custom Skills38 MCP Tools Configured

Why use Skills?

Instead of keeping an unorganized document of prompts to copy-paste, skills package expert prompts, complex multi-step workflows, and tools into a single version-controlled, portable asset. Install them once, then seamlessly invoke them with /skill-name inside your favorite AI agent. Learn more at agentskills.io →

Why Use Skills?

Portable

P

Install once Move the same workflows across Claude Code, Cursor, GitHub Copilot, Gemini CLI, Antigravity, and more.

Composable

C

Stackable Combine focused skills into richer multi-step systems without rewriting prompts every time.

Open Source

OS

Inspectable Keep your agent workflows version-controlled, reviewable, and easy to extend with your own conventions.

How these docs are built

Most pages in this site are generated from repository assets (skills/, agents/, mcp/) via wagents docs generate. The MCP overview is currently hand-maintained and summarizes the 38 servers configured in mcp.json (including local and remote servers outside this repo).

What Can You Do?

Enhance Code Reviews Research-driven code review at multiple abstraction levels with AI code smell detection and severity calibration.

Strategic Decision Analysis Domain-agnostic wargaming with Monte Carlo exploration, structured adjudication, and visual dashboards.

Host Expert Panels Simulated panel discussions among AI experts in roundtable, Oxford-style, and Socratic formats.

Create MCP Servers Build production-ready MCP servers using FastMCP v3 with tool design, testing, and deployment guidance.

Quick Install

npx skills add wyattowalsh/agents --all -g

How it Works

1. Install — npx skills add wyattowalsh/agents --all -g installs all skills globally.

2. Invoke — Type /skill-name in Claude Code, Gemini CLI, or any agentskills.io-compatible agent.

3. Enhance — Each skill provides domain expertise, structured workflows, and proven patterns that make your agent dramatically more capable.

Explore the Catalog

Skills Index Browse repository skills grouped by invocation model, with install commands and docs pages.

MCP Overview Hand-maintained overview of 38 configured MCP servers from `mcp.json`.

CLI Reference Commands for scaffolding, validation, packaging, installation, and docs generation workflows.

Featured Skills

wargame Strategic decision analysis and wargaming. Auto-classifies complexity for analysis or simulation. Use for decisions under uncertainty.

honest-review Confidence-scored code review with evidence validation. Session or full codebase audit. Use when reviewing changes or auditing quality.

skill-creator Create, improve, and audit AI agent skills. 14 structural patterns, deterministic scoring. Use when building or reviewing skills.

View all 40 skills →

MCP Servers

MCP Overview Hand-maintained overview of 38 configured MCP servers from mcp.json (local and remote).

Supported Agents

Claude Code Anthropic's official CLI for Claude.

Gemini CLI Google's command-line interface for Gemini.

Codex Autonomous coding workflows for command-line development.

Cursor The AI Code Editor.

GitHub Copilot Your AI pair programmer, right in your IDE.

Antigravity Advanced Agentic Coding assistant.

Crush Autonomous development agent focused on fast terminal workflows.

OpenCode Open coding agent with repo-native instruction support.

Resources

GitHub Source code, issues, and contributions.

CLI Reference Full command reference for the wagents CLI.

Create Your Own Build, improve, and audit skills with the skill-creator.