roveapi 1.1.0

Browser automation SDK for AI agents — Playwright-as-a-Service with accessibility trees

roveapi

Python SDK for the Rove browser automation API — Playwright-as-a-Service with accessibility trees for 77% fewer tokens.

Install

pip install roveapi

Quickstart

from roveapi import BrowserClient

client = BrowserClient(api_key="rvp_live_...")

# One-shot screenshot
screenshot = client.screenshot(url="https://example.com")
print(screenshot["url"])  # signed S3 URL

# Session with context manager (auto-close)
with client.session() as session:
    session.navigate("https://example.com")
    tree = session.get_a11y_tree()
    print(tree["result"]["tree"])  # ~26K tokens vs ~114K for screenshot

    session.click(label="Get Started")
    session.fill(selector="#email", value="[e-mail]")

Error Handling

from roveapi import BrowserClient, InsufficientCreditsError, RateLimitError

try:
    with client.session() as session:
        session.navigate("https://example.com")
except InsufficientCreditsError:
    print("Top up credits at roveapi.com")
except RateLimitError as e:
    print(f"Retry after {e.retry_after} seconds")

API

BrowserClient(api_key, base_url?, retries?, retry_delay?)

• client.session(**options) → Session (context manager)
• client.screenshot(url, **options) → dict
• client.extract(url, schema, **options) → dict
• client.usage() → dict

Session

All 12 browser actions: navigate(), get_a11y_tree(), click(), fill(), select(), scroll(), wait_for(), screenshot(), get_text(), get_attribute(), evaluate(), close().

Auto-retry on 503 with exponential backoff. No retry on 402 or 410.