Serve Mode

wavexis includes an HTTP API server powered by aiohttp. Start it with:

wavexis serve --host 0.0.0.0 --port 8080

Endpoints

Method	Path	Description
GET	/health	Health check
GET	/version	wavexis version
GET	/backends	Available backends
POST	/screenshot	Take a screenshot
POST	/pdf	Generate a PDF
POST	/eval	Evaluate JavaScript
POST	/scrape	Scrape multiple URLs
POST	/dom/get	Get DOM HTML
POST	/dom/query	Query DOM elements
POST	/navigate	Navigate to URL
POST	/har	Capture HAR data
POST	/cookies/get	Get cookies
POST	/cookies/set	Set a cookie
POST	/input/click	Click an element
POST	/input/type	Type text into an element
POST	/perf/metrics	Get performance metrics
POST	/perf/trace	Capture performance trace
POST	/cwv	Core Web Vitals scoring (LCP, CLS, INP)
POST	/modify-request	Modify requests in-flight
POST	/modify-response	Modify responses in-flight
POST	/multi	Run multiple actions
GET	/plugins	List discovered plugins

Rate limiting

Protect the HTTP API from abuse with --rate-limit:

wavexis serve --rate-limit 60

This allows up to 60 requests per minute using a token bucket algorithm. When the limit is exceeded, the server responds with 429 Too Many Requests and a Retry-After header indicating when to retry.

Backend degradation

wavexis automatically falls back from the preferred backend to an alternative if it cannot be created (e.g. cdpwave not installed). Use --backend to set a preference:

wavexis serve --backend cdp    # prefers cdp, falls back to bidi
wavexis serve --backend bidi   # prefers bidi, falls back to cdp

Core Web Vitals

Measure LCP, CLS, INP with scoring via the /cwv endpoint:

curl -X POST http://localhost:8080/cwv \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com", "budgets": {"lcp_ms": 2500}}'

Examples

Screenshot

curl -X POST http://localhost:8080/screenshot \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com"}' \
  -o screenshot.png

Eval

curl -X POST http://localhost:8080/eval \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com", "expression": "document.title"}'

Navigate

curl -X POST http://localhost:8080/navigate \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com"}'

Architecture

Each request creates a fresh backend instance via BackendManager.select_with_fallback(). If the preferred backend fails to initialize, wavexis automatically tries the next available backend. The _run_action helper wraps action.execute() + close() in a try/finally block, ensuring the browser is always cleaned up.

WebSocket streaming

The /ws endpoint provides real-time streaming of browser events via WebSocket.

Subscribe

Connect to ws://localhost:8080/ws and send a JSON subscribe message:

{
    "url": "https://example.com",
    "events": ["screenshot", "console", "navigation", "dom_mutation", "perf_metrics"],
    "interval": 1.0,
    "format": "png",
    "quality": 80
}

Streamed events

The server sends JSON messages with a type field:

Type	Description
ready	Server is ready, browser launched and navigated
screenshot	Base64-encoded screenshot image
console	Console message (deduplicated)
navigation	URL change detected
dom_mutation	DOM content change detected (polled)
perf_metrics	Performance metrics snapshot (LCP, FCP, CLS, etc.)
network_request	Network request intercepted
network_response	Network response intercepted
dialog	JavaScript dialog event
navigated	Confirmation of client-initiated navigation
eval_result	Result of a client-initiated eval
error	Error from a stream source

Client commands

Send JSON commands to control the browser:

{"action": "navigate", "url": "https://example.org"}
{"action": "eval", "expression": "document.title"}
{"action": "screenshot"}
{"action": "close"}

Available event types

Event	Description	Streaming method
screenshot	Periodic screenshots (base64)	Polling
console	Console messages (deduplicated)	Polling
navigation	URL changes	Polling
dom_mutation	DOM content changes	Polling
perf_metrics	Performance metrics snapshots	Polling
network_request	Network requests	Event subscription
network_response	Network responses	Event subscription
dialog	JavaScript dialogs	Event subscription

Python client example

import asyncio
import json
import aiohttp

async def stream():
    async with aiohttp.ClientSession() as session:
        async with session.ws_connect("ws://localhost:8080/ws") as ws:
            await ws.send_json({
                "url": "https://example.com",
                "events": ["screenshot", "console"],
                "interval": 2.0,
            })
            async for msg in ws:
                if msg.type == aiohttp.WSMsgType.TEXT:
                    data = json.loads(msg.data)
                    if data["type"] == "screenshot":
                        print(f"Got screenshot ({len(data['data'])} bytes)")
                    elif data["type"] == "console":
                        print(f"Console: {data['data']}")
                    elif data["type"] == "ready":
                        print(f"Ready: {data['url']}")

asyncio.run(stream())

JavaScript client example

const ws = new WebSocket("ws://localhost:8080/ws");
ws.onopen = () => {
    ws.send(JSON.stringify({
        url: "https://example.com",
        events: ["screenshot", "console", "navigation"],
        interval: 1.0,
    }));
};
ws.onmessage = (event) => {
    const data = JSON.parse(event.data);
    if (data.type === "screenshot") {
        console.log(`Screenshot: ${data.data.length} chars`);
    } else if (data.type === "console") {
        console.log("Console:", data.data);
    } else if (data.type === "navigation") {
        console.log("Navigated to:", data.url);
    }
};