Give Your AI Agent Eyes on the Web

{
  "mcpServers": {
    "rendex": {
      "command": "npx",
      "args": ["-y", "@copperline/rendex-mcp"],
      "env": {
        "RENDEX_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

# Using the Rendex MCP server with Claude
# After configuring MCP, your agent can simply ask:
# "Take a screenshot of https://example.com"
# "Generate a PDF of this webpage"
# "Render this HTML to an image"

# Or use the SDK directly in agent code:
from rendex import Rendex

rendex = Rendex("YOUR_API_KEY")

# Agent captures a deployment for verification
result = rendex.screenshot(
    url="https://staging.myapp.com",
    full_page=True,
    format="png",
)

# Agent analyzes the result
print(f"Page loaded in {result.metadata.load_time_ms}ms")
print(f"Captured {result.metadata.bytes_size} bytes ({result.metadata.format})")

# Remote MCP server — no npm install needed
# Add to your MCP client config:
# Server URL: https://mcp.rendex.dev/mcp
# Auth: Bearer YOUR_API_KEY

# Or use the stdio transport locally:
npx -y @copperline/rendex-mcp

# The MCP server exposes these tools:
# - rendex_screenshot: Render a URL, HTML, or Markdown to an image or PDF
# - rendex_extract: Pull clean reader-mode Markdown/JSON/HTML from a URL