Site auditing powered by Plasmate - 10x faster than Chrome-based tools.
pip install plasmate plasmate-audit# Audit a site (crawls up to 50 pages)
plasmate-audit https://example.com
# More pages
plasmate-audit https://example.com --max-pages 200
# JSON output
plasmate-audit https://example.com --jsonfrom plasmate_audit import audit_site
result = audit_site("https://example.com", max_pages=50)
print(f"Score: {result['score']}/100")
print(f"Pages: {result['pages_crawled']}")
print(f"Issues: {result['errors']} errors, {result['warnings']} warnings")| Check | Severity | Description |
|---|---|---|
| missing-title | Error | Page has no title |
| title-too-long | Warning | Title >60 characters |
| title-too-short | Warning | Title <10 characters |
| no-h1 | Error | No H1 heading |
| multiple-h1 | Warning | More than one H1 |
| no-headings | Warning | No headings at all |
| empty-links | Warning | Links with no text |
| thin-content | Warning | Pages with <100 chars of content |
| images-no-alt | Error | Images without alt text |
Screaming Frog and Lighthouse use Chrome to render every page. Plasmate skips rendering:
| Tool | Speed (100 pages) | Memory |
|---|---|---|
| Screaming Frog | ~5 minutes | 2GB+ |
| Lighthouse | ~15 minutes | 500MB+ |
| plasmate-audit | ~30 seconds | 50MB |
- Crawl — Start from a URL, discover internal links via Plasmate's SOM
- Analyze — Run SEO/content checks against each page's structured object model
- Report — Get a score, issue list, and per-page breakdown
No headless browser. No JavaScript execution. Just fast, structured analysis.
Apache 2.0
| Engine | plasmate - The browser engine for agents |
| MCP | plasmate-mcp - Claude Code, Cursor, Windsurf |
| Extension | plasmate-extension - Chrome cookie export |
| SDKs | Python / Node.js / Go / Rust |
| Frameworks | LangChain / CrewAI / AutoGen / Smolagents |
| Tools | Scrapy / Audit / A11y / GitHub Action |
| Resources | Awesome Plasmate / Notebooks / Benchmarks |
| Docs | docs.plasmate.app |
| W3C | Web Content Browser for AI Agents |