@agentlee5/agent-skills

LeeWay Agent Skills

The developer-facing skill system for agentic engineering.

LeeWay Agent Skills gives developers, AI engineers, platform teams, and product builders a practical way to turn LLM capability into an operating layer: reusable skill modules, callable MCP tools, workflow orchestration, governance automation, and proof-backed integration assets.

Why This Project Exists

Most AI tooling stops at prompting. LeeWay Agent Skills is built for teams that need more than a clever demo. It packages structured execution knowledge, workflow composition, MCP delivery, and compliance controls into one repository so agentic systems can be deployed with more consistency, more speed, and less operational guesswork.

If you are building internal copilots, engineering assistants, AI-enabled products, platform tooling, or multi-agent systems, this project is designed to help you move from experimentation to repeatable delivery.

What You Get

Area	What it delivers
Skill Library	51 installed skill modules across 27 categories for engineering, QA, AI, security, architecture, workflow composition, and more
MCP Runtime	44 active MCP-exposed tools today through the packaged skills registry
Workflow Layer	50+ documented workflow patterns and acquisition plans for orchestration at scale
Standards and Governance	Leeway Standards integration for headers, compliance scanning, remediation, and operational guardrails
NPM Distribution	Installable package and CLI surface via @leeway/agent-skills
Integration Proof	Badge bundle generator to visibly prove another application is powered by LeeWay Agent Skills

Built For

• Developers building faster with structured AI capabilities instead of ad hoc prompting
• Engineering organizations that want reusable agent skills with governance built in
• AI product teams shipping assistants, copilots, internal platforms, or MCP servers
• Agencies, consultants, and platform teams who need proof-backed AI integration assets
• Multi-agent and workflow builders who need a growing skill and orchestration layer

Product Positioning

LeeWay Agent Skills is not just a prompt collection and not just a documentation dump. It is a packaged capability layer for agentic software systems.

It combines:

• A reusable skill library
• A callable MCP tool surface
• Workflow composition strategy
• Governance and compliance automation
• Badge-driven proof of integration
• A distribution model that works through GitHub and npm

That combination makes it useful both as an engineering asset and as a platform story you can deploy, demonstrate, and sell internally or externally.

Quick Start

Install From npm

npm install @leeway/agent-skills

leeway-agent-skills info
leeway-agent-skills-mcp
leeway-agent-skills-badge .\public\leeway-proof "My App" mcp-server

Compatibility aliases are also included:

leeway-skills info
leeway-skills-mcp

Clone The Repository

git clone https://github.com/4citeB4U/LeeWay-Agent-Skills.git
cd LeeWay-Agent-Skills
npm --prefix mcp-server install
npm --prefix mcp-server run build
node .\bin\leeway-skills.js info

Run The MCP Server

node .\bin\leeway-skills.js mcp

Generate A Proof-Backed Badge Bundle

node .\bin\leeway-skills.js badge .\public\leeway-proof "My App" mcp-server

How Developers Use It

Use case	What LeeWay Agent Skills adds
Internal engineering copilots	Reusable skill modules, clearer execution prompts, and callable tools
CI and delivery automation	Workflow composition patterns for diagnosis, recovery, validation, and release handling
Code generation and review	Structured expertise across TypeScript, Python, testing, architecture, and security
Security and compliance	Standards checks, header enforcement, secret scanning, and remediation helpers
Productized AI platforms	Badge proof bundles and npm distribution for integration into other apps

Architecture At A Glance

Skills -> Workflows -> MCP Tools -> Agents -> Governance -> Proof

LeeWay Agent Skills
  |- skills/                 reusable capability modules
  |- mcp-server/             packaged Model Context Protocol server
  |- LeeWay-Standards/       governance, scanning, and repair tooling
  |- scripts/                automation and registry utilities
  |- documents/              strategy, planning, architecture, and rollout docs
  |- bin/                    package-level developer CLIs
  `- sdk/                    importable runtime helpers and path utilities

Core Capabilities

Skill Library

The skills/ tree contains the actual operating knowledge for the system. These skills cover code generation, debugging, testing, infrastructure, AI/ML, architecture, security, research, workflow composition, and more.

MCP Server

The packaged MCP server exposes the active skill registry as callable tools so LLMs and agent frameworks can interact with the system as runtime capabilities instead of static reference text.

Start here: mcp-server/README.md

Workflow Strategy

LeeWay Agent Skills includes workflow planning and orchestration strategy for higher-order execution, including pre-composed bundles for CI recovery, reporting, coordination, analysis, and agent collaboration.

Start here: documents/LEEWAY_WORKFLOWS_STRATEGIC_PLAN.md

Governance And Standards

The project integrates Leeway Standards so teams can add headers, scan compliance posture, and improve consistency across files, agents, and automation pipelines.

Start here: documents/LEEWAY_STANDARDS_COMPLIANCE.md

Badge And Proof Integration

If another application uses LeeWay Agent Skills, the badge tooling can generate visible proof files and embeddable assets so usage is not hidden or hand-wavy.

Start here: documents/LEEWAY_BADGE_INTEGRATION.md

Repository Guide

Path	Purpose
README.md	Project overview, positioning, and developer onboarding
package.json	Root npm package for @leeway/agent-skills
skills/	Installed skill library
mcp-server/	MCP runtime source and packaged build output
LeeWay-Standards/	Governance toolkit and CLI
scripts/	Registry, compliance, bootstrap, and workflow tooling
documents/	Strategic, technical, and operational documentation
sdk/	Importable helpers for packaged consumers
bin/	CLI entrypoints for package users

Documentation Map

• NPM SDK Guide
• MCP Architecture
• Agent Lee Integration
• Leeway Standards Compliance
• Workflow Strategic Plan
• Skill Acquisition Executive Summary

Current Status

Signal	Status
Repository path	C:\Tools\LeeWay-AgentSkills
Package name	@leeway/agent-skills
Installed skill modules	51
Installed categories	27
Active MCP tools in registry	44
Workflow roadmap	50+ documented patterns
Expansion roadmap	250+ target skill acquisition plan

Roadmap

• Sync the active registry to fully reflect all installed skills
• Publish and stabilize the npm package lifecycle
• Expand toward the 250+ skill acquisition roadmap
• Add more production-ready workflow bundles and integration examples
• Strengthen productized deployment stories for internal AI platforms and external SaaS products

Why It Matters For Engineering Leaders

LeeWay Agent Skills gives technical teams a cleaner answer to a hard problem: how do you make AI behavior more reusable, inspectable, composable, and governable?

This repository answers that with a system developers can actually adopt:

• installable
• inspectable
• documentable
• extensible
• compatible with MCP
• aligned with governance

That makes it easier to turn AI from a fragile experiment into an engineering capability.

Contributing

Contributions should preserve the structure of the skill system, keep non-README markdown inside documents/, and maintain compatibility with the Leeway Standards governance model.

License

MIT. See LICENSE.