Fleet Generator
A single-file, browser-based wizard that generates a ready-to-use multi-agent Claude Code / Codex setup — orchestrator, workers, and a reviewer.
🔗 Visit Fleet Generator
Description
Setting up a team of AI coding agents that actually divide work sensibly — one agent planning, several doing focused work, one checking quality — usually means hand-writing a bunch of configuration files and figuring out which agent should use which model tier to avoid burning your budget on a model that's overkill for simple tasks. Fleet Generator turns that setup process into a short questionnaire you answer in your browser.
Fleet Generator is a self-contained HTML file — no installation, no build step, no server — that you open locally and click through a configuration wizard. It generates an orchestrator agent (which breaks work into tasks and assigns them), a set of worker agents, and a dedicated reviewer agent for quality control, assigning appropriately-sized models to each role to control cost (the project claims roughly 64% lower token cost versus a single do-everything agent). The output is a ready-to-use `.claude/` directory bundle for Claude Code, with an optional `AGENTS.md` for Codex compatibility and optional integration with a persistent-memory tool called MOOTx01. Everything runs locally in the browser with no external API calls during setup.
💬 Our review
The short version: if you've been meaning to set up a proper multi-agent Claude Code fleet (orchestrator + workers + reviewer) but keep putting it off because hand-writing the configuration is tedious, this collapses that setup into a five-minute questionnaire — the cost claim (64% token savings) is the vendor's own number and worth verifying on your own workload rather than taking at face value.
There's no direct competitor doing exactly “generate a fleet config from a wizard,” since most people either copy a template repo or write the orchestrator/worker/reviewer structure by hand from documentation. Fleet Generator's real value is removing that initial setup friction, especially the non-obvious part of assigning cheaper models to worker roles and reserving stronger models for orchestration and review. Because it's a single static HTML file with no server, there's very little to trust or audit beyond reading the file yourself, which is reassuring for a tool that's about to configure how your coding agents behave. Worth trying if you're about to set up a multi-agent Claude Code or Codex workflow from scratch; less useful if you already have a working fleet configuration you're happy with.
💰 Pricing
📊 Global score
🤖 AI-enriched data
Gratuit, open source (MIT), aucune installation requise.
Pros
Assistant en 5 minutes au lieu d'une config manuelle
Aucune installation, fichier HTML unique, zéro appel API externe
Attribution automatique de modèles adaptés au coût par rôle
Compatible Claude Code et Codex
Cons
Le chiffre de -64% de coût token est une affirmation du projet, non vérifiée indépendamment
Utile seulement à la mise en place initiale
Projet jeune (6 étoiles), config générée à relire avant usage