Directory
‹ Back to apps

Lugh

Write, check and dry-run AI agent programs in Lugh, where permissions, budgets and human approval are part of the code. Every action passes a policy gate and each run ends with a policy log.

What’s inside

1 MCP server · 1 skill
  • SkilllughWrite, check and dry-run AI-agent programs in Lugh, where permissions, budgets and human approval are part of the code. Use when a person wants an agent that touches real systems (files, HTTP APIs, shell, MCP tools, payments, email) and wants to see and limit what it will do before it runs.
  • MCP serverlughMCP server with tools to read the Lugh reference, check a program, dry-run it in a sandbox, run its tests and describe the sandbox. All tools are read-only.