Floway One — AI skill for Claude Code
Floway One — your AI model gateway, running locally on your machine, just like Clash.
How to install Floway One
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open tommy0103/Floway-One and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Floway One does
Floway One — your AI model gateway, running locally on your machine, just like Clash.
Alternatives in AI
- Ante — Ghost in your shell 1.9k ★
- ML Training Pack — Machine learning model training skills 1.6k ★
- Nocturne Memory — A lightweight, rollbackable, and visual Long-Term Memory Server for MCP Agents 1.3k ★
README
Floway
All models, one local entry.
Floway is a personal, local-first LLM gateway for people who use multiple AI coding tools. It puts subscription-backed and token-backed model providers behind one stable loopback address, then routes each model through the API shape the client already speaks — like Clash, but for AI model traffic.
This repository is a personal desktop edition of [Floway](https://github.com/Menci/Floway). It keeps the full gateway, protocol translation, provider integrations, and Dashboard, and packages them as an installable desktop application that runs entirely on your machine. No Docker, Node.js, or database to install; no account or cloud control plane to trust. The product and technical specification lives in [docs/floway-one-spec.zh-CN.md](./docs/floway-one-spec.zh-CN.md).
Highlights
- One stable address: every client always talks to
http://127.0.0.1:8788. Configure Codex, Claude Code, and other tools once; switch providers, models, and routing afterwards without touching client configuration. - Use GitHub Copilot, ChatGPT subscriptions, Claude.ai subscriptions, Azure AI, configurable multi-protocol HTTP providers, and Ollama from one app.
- Serve OpenAI, Anthropic, Gemini-compatible, audio transcription, and rerank APIs with cross-protocol translation where needed.
- One owner, many API keys: give each tool, project, or use case its own revocable key with per-key upstream scope, usage statistics, and request history.
- Local-first: configuration, credentials (encrypted at rest), usage, and request records stay in the operating system's per-user application-data directory.
- Desktop shell with tray lifecycle: closing the window keeps the gateway running; only an explicit quit stops it.
- Encrypted full backups and credential-free safe exports.
- Generate one-command Claude Code and Codex configurations from an API key.
Quick Start
Signed desktop installers are on the roadmap; today Floway runs from source.
Related Skills
Floway
Serverless LLM API gateway with telemetry, full protocol translation and Stateful Responses support. Use any L
Which Model
Which model actually answered your coding agent, from your own machine: Claude Code and Codex transcripts read
Fallback
Show or change hmd's OmniRoute fallback-routing gate — whether Heimdall may route work to a third-party model,
Use Crew Gateway
Route an OpenAI/Anthropic-compatible client through the portable farmd control plane's virtual crew model, for
Client
Run Previously locally — an AI agent that remembers by when, not by chat thread. One npm package: the same ker
Harness Scouter
Scores a Claude Code harness from the transcripts already on your machine: 6 stats of what your agent actually
Related Agents
Privacy Policy
Last updated: 2026-02-28 nWave is an open-source AI workflow framework that runs inside Claude Code and stores
Lmstudio Local
A locally-hosted model served through LM Studio's OpenAI-compatible API,
GPT 5.6 Sol
General-purpose agent that runs on GPT-5.6-Sol (routed through the shunt gateway to the ChatGPT/Codex subscrip