🐒 Apeeye

Apeeye is a zero-bullshit mock API tool. Create fake REST endpoints in seconds. No setup. No login. No database.

• Define collections like /users or /posts and get a live REST URL instantly
• Full GET, POST, PUT, PATCH, DELETE with sensible status codes
• Your data never leaves your browser. The server stores nothing.
• Call your endpoints from your app, curl, Postman, or a teammate's machine
• Export and import your whole workspace as JSON

Demo

Apeeye-vid-hq.mp4

How it works

Your browser mints a random, unguessable workspace ID on first visit and stores all your collections and records in IndexedDB. A tiny stateless relay server makes them reachable at a real URL:

https://<host>/mock/<workspace>/users

When anything calls that URL, the relay forwards the request over a WebSocket to your open tab, the tab answers from IndexedDB, and the relay returns the response. Nothing is ever written on the server.

The one trade-off: endpoints answer only while a tab with your workspace is open. Close it and callers get a clear 503. Reopen it and everything is back.

Run it locally

git clone https://github.com/ibr0r0/apeeye.git
cd apeeye
npm install

Option A — the way production runs:

npm run build:web        # builds the frontend into dist/
npm run relay            # wrangler dev: serves dist/ and the /mock endpoints

Open http://localhost:8787.

Option B — hot-reloading frontend during development:

# terminal 1
npm run relay

# terminal 2
npm run web              # Expo dev server on :8081, talks to the relay on :8787

The frontend finds the relay automatically when served by it. When running Expo separately, it defaults to http://localhost:8787; override with EXPO_PUBLIC_RELAY_URL (see .env.example).

Usage

1. Open the Playground. Your workspace URL is shown at the top; copy it.
2. Add a collection (e.g. users). It is a live endpoint immediately.
3. Add records with the field builder or paste raw JSON. Or just POST to the URL.
4. Point your app at https://<host>/mock/<workspace>/users and fetch.
5. Watch calls arrive in the live Requests feed.

const res = await fetch("https://<host>/mock/<workspace>/users");
const users = await res.json();

Endpoint reference, examples and limits are in the in-app Docs tab.

Security

There are no accounts. The workspace ID (10 chars, base62, CSPRNG) is a capability: anyone with the URL can read and write that workspace. Reset rotates it.

The relay is hardened for public exposure:

• Strict CSP, HSTS, nosniff, frame denial and a same-origin check on the WebSocket endpoint
• Per-IP and per-workspace rate limits, body size cap, relayed-frame cap, max in-flight requests per workspace, unregistered sockets dropped in 5 s
• A tab can only answer requests for its own workspace; forged replies are ignored
• Dangerous response headers from tabs are stripped (set-cookie, CSP, HSTS…)
• Prototype-polluting keys are stripped from every payload
• Nothing is persisted server-side, so there is nothing to leak

This is a mocking tool. Do not put real user data in it.

Tests

npm test

Unit tests for the shared mock engine, wire chunking, response hygiene, workspace ID and import validation, plus integration tests that boot the real Worker under wrangler dev and drive it with a fake tab over WebSocket. Uses Node's built-in runner.

Tech Stack

• Frontend: React Native Web via Expo, IndexedDB, WebSocket
• Relay: Cloudflare Workers + Durable Objects, stateless
• Shared: a storage-agnostic mock engine used by both the UI and relayed requests

Roadmap

• Local mock API with a web UI
• Browser-side storage, per-visitor workspaces, stateless relay
• Shareable endpoints with unique URLs
• Import/export workspaces as JSON
• Live request feed
• Delay and error simulation (e.g. 404, 500, timeout)
• Testing presets for frontend devs (auth flow, search, pagination)
• Optional Service Worker mode for fully offline in-page mocking

Contributions are welcome. Open an issue or a PR for anything on the list.

Contributing

• Keep it simple and clean.
• Run npm test and make sure the app still works locally.
• If you add a feature, explain it clearly in your PR.

Thanks for helping improve Apeeye 🙌

---

• 
• 
• 
• 

---

For support, suggestions, or general inquiries, please reach out via X at @ibr0r.