These docs cover everything you need to get up and running.
| Document | Description |
|---|---|
| Quick Start | Docker setup, first request in 10 minutes |
| Integration Guide | Point Python, Node.js, or curl clients at the proxy |
| Feature Walkthrough | Cost attribution, budgets, alerts, runaway detection |
| FAQ & Troubleshooting | Common issues, auth, streaming, supported models |
Your LLM Client
│ base_url → http://localhost:3001/openai
│ X-SteadIO-Key, X-Agent-Id headers
▼
SteadIO Proxy (3001)
│ Auth check → Budget check → Forward to provider
│ Fire-and-forget cost event
▼
Cost Engine (3002)
│ Calculate cost → Budget enforce → Runaway detect
│ Store in PostgreSQL + Redis
▼
Dashboard (5173)
│ Cost by agent/team, budgets, runaway history
| Service | Endpoint | Purpose |
|---|---|---|
| Proxy | http://localhost:3001/openai/* |
OpenAI-compatible proxy |
| Proxy | http://localhost:3001/anthropic/* |
Anthropic proxy |
| Cost Engine | http://localhost:3002/api/budgets |
Budget CRUD |
| Cost Engine | http://localhost:3002/api/runaways |
Runaway event history |
| Dashboard | http://localhost:5173 |
Web UI |