@shift-zero/omni-config
v1.0.0
Published
Configuration toolkit for OmniRoute — generate AI tool configs from one CLI. Hermes, Claude Code, Codex, Cline.
Maintainers
Readme
🛸 omni-config
Configuration toolkit for OmniRoute — the free AI gateway with 271+ providers, 90+ free tiers, auto-fallback, and RTK+Caveman compression.
Generate, test, and manage OmniRoute configuration for all your AI coding tools from one CLI.
npm install -g omni-configWhy?
OmniRoute is incredible — one endpoint, auto-fallback across 271 providers, saves 15-95% on tokens. But configuring every AI coding tool to use it is tedious. omni-config automates that:
- One command per tool — generate ready-to-use config files
- Doctor — check if OmniRoute is running and which tools are configured
- Init wizard — set up everything interactively
Usage
# Check if OmniRoute is running
omni-config test
# Generate config for a specific tool
omni-config generate hermes # Hermes Agent
omni-config generate claude-code # Claude Code
omni-config generate codex-cli # Codex CLI
omni-config generate cline # Cline VS Code extension
omni-config generate generic # Generic OpenAI-compatible env vars
# Full diagnostic
omni-config doctor
# Interactive setup wizard
omni-config init
# List supported tools and their config paths
omni-config list
# Fetch OmniRoute provider catalog
omni-config providersOptions
omni-config generate hermes --url=http://192.168.1.5:20128 --key=sk-xxx --model=auto/coding
omni-config generate generic --stdout # Print to stdout
omni-config generate codex-cli --output=./config.jsonSupported Tools
| Tool | Command | Config File |
|------|---------|-------------|
| 🛸 Hermes Agent | generate hermes | ~/.hermes/.env |
| 🤖 Claude Code | generate claude-code | ~/.claude/claude.json |
| 🔮 Codex CLI | generate codex-cli | ~/.codex/config.json |
| 🧩 Cline | generate cline | ~/.config/cline/config.json |
| 🔗 Generic | generate generic | stdout or custom path |
What is OmniRoute?
OmniRoute is a free, MIT-licensed AI gateway that gives you:
- 271 AI providers (90+ free tiers) through one OpenAI-compatible endpoint
- Auto-fallback — quota runs out? Slides to the next provider silently
- RTK + Caveman compression — saves 15-95% on tokens (~89% average)
- 18 routing strategies — priority, cost-optimized, fusion, pipeline, etc.
- Built-in MCP server (104 tools) + A2A agent protocol
- ~1.4B free tokens/month from documented provider free tiers
npm install -g omniroute
omniroute # Starts the gateway on http://localhost:20128License
MIT — like the project it helps configure.
Built by Zero 🛸 for Jacob.
