# Gatewayz Wiki

**Gatewayz** is a universal AI inference gateway — one API that connects to 30+ AI provider gateways, giving access to 10,000+ models. Developers integrate once, get everything.

> **One API key, every AI model, automatic reliability, one bill.**

---

## New Here? Start with the Reading Path

These 6 docs tell the full story of Gatewayz — from "what is it" to "what still needs work." Read them in order. Each one builds on the last.

| # | Document | What you'll learn | Time |
|---|----------|------------------|------|
| 1 | **[[Conceptual Model]]** | What Gatewayz is, the architecture, all 10 system layers | 15 min |
| 2 | **[[Stability Definition]]** | What "stable" means from user, product, and engineering perspectives | 5 min |
| 3 | **[[Conceptual Model Features]]** | All 56 features the system must have (the spec) | 20 min |
| 4 | **[[Features]]** | What's actually built today — 450+ endpoints across 17 systems | 25 min |
| 5 | **[[Delta Report]]** | Spec vs reality: what's done, what's missing, what to fix (P0/P1/P2) | 15 min |
| 6 | **[[Features-Acceptance-Criteria]]** | How we verify each feature is correctly implemented | 20 min |

> After reading these 6 docs, you'll understand what Gatewayz is, what it should do, what it currently does, where the gaps are, and how we measure "done."

---

## Testing & Verification

Once you understand the system, these docs cover how we test it.

| Document | What it covers |
|----------|---------------|
| **[[Testing Guide]]** | Start here — overview of all testing docs and how to run tests |
| **[[Testing Plan]]** | Manual API test cases (curl/Postman) for every feature |
| **[[CM Unit Testing Plan]]** | Unit tests derived from conceptual model claims |
| **[[CM Unit Test Coverage Report]]** | Which unit tests exist vs missing (196 dedicated tests, 5 xfail gaps) |
| **[[Test Coverage Audit]]** | Feature-by-feature automated test coverage |

---

## System Reference

Deep dives into individual systems. Read these when you need to understand or modify a specific area.

### Security & Access
- [[API-Key-Management]] — Key encryption, scopes, IP allowlists, rotation
- [[Role-Based-Access-Control]] — Roles, permissions, hierarchy
- [[Activity-Logging]] — User action and system event logging
- [[Audit-System]] — Security audit trail, tamper protection

### Billing & Monetization
- [[Stripe-Integration]] — Payments, webhooks, credit calculation
- [[Subscription-Plans]] — Plan tiers, lifecycle, credit allocation
- [[Free-Trial-System]] — Trial limits, lifecycle, conversion
- [[Coupon-System]] — Coupon creation, redemption, validation
- [[Referral-System]] — Referral codes, bonus rules, webhook flow

### Monitoring & Observability
- [[Monitoring-System]] — Endpoints, Redis/Postgres data, circuit breakers
- [[Performance-Monitoring]] — TTFB, streaming duration, Prometheus histograms
- [[Prometheus-Setup]] — Metric categories, Railway deployment, PromQL
- [[Error-Monitoring]] — Sentry integration, alerts, performance spans

### Features
- [[Chat-History]] — Persistent sessions, history injection
- [[Feature-Flags]] — Statsig gates, configs, experiments

### Provider Integrations
- [[Alibaba-Integration]] — DashScope / Qwen model setup
- [[Google-Vertex-Setup]] — Gemini models, auth methods, regional endpoints

### Operations
- [[Supabase-Migrations-CI]] — GitHub Actions CI/CD for migrations
- [[Testing-Workflows-Locally]] — Running GitHub Actions locally with `act`
- [[Troubleshooting]] — Common issues and resolutions

---

## Data References

These are large generated documents — useful for lookup, not for reading start to finish.

- [[API-Mappings]] — Complete endpoint-to-code mapping (450+ endpoints)
- [[Test-Mapping]] — Test-to-feature mapping (~5,268 tests across 334 files)

---

*Version 2.0.4 | Last updated: March 2026*