mcp

Arkestone MCP Servers

Overview

A monorepo hosting Model Context Protocol (MCP) servers for GitHub Copilot customization. Each server implements the MCP specification to dynamically serve different types of Copilot configuration — custom instructions, skills, and more — from local directories and GitHub repositories.

Servers

Shared Components

• pkg/optimizer — OpenAI-compatible LLM optimization layer shared by all servers. Optionally merges, deduplicates, and consolidates content from multiple sources via any OpenAI-compatible endpoint.
• pkg/config — Unified configuration loading (YAML → env vars → CLI flags).
• pkg/github — GitHub Contents API client with proxy and header pass-through.
• pkg/httputil — Shared HTTP infrastructure: proxy support, custom TLS/CA, header propagation.
• pkg/syncer — Background periodic sync for remote repositories.
• pkg/server — MCP server bootstrap helpers.

Quick Start

mcp-instructions

make build-instructions
./bin/mcp-instructions -dirs /path/to/repo

mcp-skills

make build-skills
./bin/mcp-skills -repos github/awesome-copilot

See each server's README for full configuration and usage details.

Architecture

.
├── servers/
│   ├── mcp-instructions/       # mcp-instructions server
│   │   ├── cmd/mcp-instructions/
│   │   └── internal/loader/
│   └── mcp-skills/             # mcp-skills server
│       ├── cmd/mcp-skills/
│       └── internal/scanner/
├── pkg/
│   ├── config/             # shared configuration loading
│   ├── github/             # GitHub Contents API client
│   ├── httputil/           # proxy, TLS, header propagation
│   ├── optimizer/          # shared LLM optimization layer
│   ├── server/             # MCP server helpers
│   └── syncer/             # background sync
├── docs/
│   └── network.md          # network / proxy / firewall guide
├── examples/               # client configuration examples
├── AGENTS.md               # AI coding assistant guide
├── Makefile
├── go.mod
└── go.sum

Each server follows the same layered design:

1. Config — YAML → environment variables → CLI flags (each layer overrides the previous)
2. Loader / Scanner — discovers content from local directories and GitHub repositories
3. Optimizer — optional LLM-based consolidation via pkg/optimizer
4. MCP Server — exposes content as Resources, Prompts, and Tools over stdio or Streamable HTTP

Development

make build              # build all servers
make test               # run unit tests
make test-integration   # run integration tests
make docker             # build Docker images for all servers
make lint               # run golangci-lint
make cover              # generate coverage report

Configuration

Each server supports configuration via YAML files, environment variables, and CLI flags. See the individual server READMEs for details:

• mcp-instructions configuration
• mcp-skills configuration

GitHub Authentication (Optional)

A GitHub token is optional. Public repositories work without authentication. For private repositories, provide a token via any of these methods (highest priority first):

When a repository returns HTTP 401/403/404 without a token configured, the error message will hint that authentication may be required.

Network & Proxy

All servers work on-premise, in private/public cloud, with direct internet or through HTTP/HTTPS proxies. See the Network & Proxy Guide for:

• Firewall rules to open
• Proxy and custom CA certificate configuration
• HTTP header pass-through from incoming requests to outbound calls
• Deployment scenarios (direct, proxy, air-gapped, Kubernetes)

Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines.

License

This project is licensed under the MIT License — see the LICENSE file for details.