Why GoDizzy
AI agents depend on external tools: search APIs, payment services, LLM providers, internal microservices. Those tools fail, rate-limit, drift, and behave unpredictably in ways that are hard to reproduce. GoDizzy gives you deterministic control over every tool call without modifying a single line of agent code.One stable URL
Each route collection gets a permanent subdomain on
godizzy.dev. Your agent always calls the same address — you change the behavior behind it.Mock or proxy per route
Set each routing rule to return a fixture or forward to a live backend. Toggle between them in the dashboard, route by route.
Latency injection
Add a min/max delay range to any mock rule. Simulate slow responses and test that your agent handles timeouts and retries correctly.
No agent code changes
Switch from mock to proxy, change a fixture, or flip environments — none of it requires redeploying or editing your agent.
Key features
Routing rules — Each rule matches on HTTP method and path pattern. Rules are evaluated in priority order; the first match wins. A default fallback rule catches anything that doesn’t match a specific rule. Mock responses — Return deterministic JSON fixtures with a configured status code, headers, and optional latency range. Mock responses are versioned so you can roll back to a previous fixture at any time. Proxy forwarding — Forward requests to a real backend URL. Optionally reshape the upstream response — override status, headers, or JSON body fields — without touching the client. Environments — Separate routing configurations for production, staging, and per-developer sandboxes. Same URL format across all of them; each environment’s rules are completely isolated. Outbound webhooks — Attach HTTP callbacks to mock rules. GoDizzy fires them after serving the mock response, with configurable delays and templated payloads.Who uses GoDizzy
AI agent builders use GoDizzy to test how their agents behave when tools return errors, partial results, or unexpected schemas — before those conditions ever appear in production. Frontend developers use it to unblock UI work when backend APIs aren’t ready, serving realistic fixtures without any special backend deploy. QA and staging teams use it to force specific failure modes — 429s, 500s, timeouts, malformed payloads — on demand, reproducibly, in any environment.Get started
Quick Start
Create a gateway and route your first request in a few minutes.
How It Works
Understand the routing model: environments, collections, rules, and request flow.