LightsOut
Agent orchestration system that runs coding-agent sessions (Claude Code, Codex) with policy mediation and exposes tools via MCP.
README
LightsOut
Agent orchestration system: it runs chains of coding-agent sessions (Claude Code, Codex) inside one container, mediates their permission requests against policy, raises doubts instead of guessing, asks a second opinion when a doubt is resolvable, and exposes the whole thing over MCP plus a read-only web panel.
Copyright 2026 Filipe Gomes. Licensed under the Apache
License, Version 2.0 with the Commons Clause condition (see LICENSE): free to use, modify and
redistribute, including commercially inside your own organisation; not for sale or resale, as
itself or as a service, by anyone but the Licensor.
Documents
doc/INSTALL.md— full installation guide, including a fresh machine.doc/EXAMPLES.md— realistic usage walkthroughs.doc/REQUIREMENTS.md— what the system must do (requirement IDs).doc/DESIGN.md— how it is built, section by section.doc/STATE.md— current phase, last milestone, next step.doc/DECISIONS.md— non-obvious implementation choices.doc/TESTING.md— what can be exercised today and how.
Requirements
- Windows 11 or macOS with Docker Desktop, or Linux with Docker Engine + compose v2.
- A Claude subscription or API key, and a ChatGPT/OpenAI subscription or API key.
Install
docker run -d --name lightsout --restart unless-stopped \
-p 127.0.0.1:8484:8484 -p 127.0.0.1:1455:1455 \
-v lightsout-db:/data \
-v "$HOME/LightsOut:/workspace" \
-v claude-auth:/home/app/.claude -v codex-auth:/home/app/.codex \
ghcr.io/<owner>/lightsout:latest
Open http://127.0.0.1:8484/ — the first-run wizard walks through connecting the engines,
Claude Desktop and your first project. Install dist/lightsout.mcpb in Claude Desktop to get the
tools (see extension/README.md).
Panel and API are bound to localhost only. Health: curl -s localhost:8484/health.
Platform-specific detail — Windows double-click entry points, building from source instead of
pulling, the egress allowlist, updating — is in doc/INSTALL.md.
Egress allowlist (RT-05)
By default outbound traffic is unrestricted and /health says so
(network: unrestricted). To enforce the allowlist, edit proxy/filter with your
git remote hosts and start with the overlay:
docker compose -f docker-compose.yml -f docker-compose.secure.yml --profile secure up -d
Development
npm install
npm run typecheck
npm run build
npm test
The build order and the per-phase verification scripts are defined in doc/DESIGN.md §13; each
phase has a gate under scripts/verify/. The gates need a shell with docker on PATH: on
Windows, Git Bash works.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.