# How crawwwl works

crawwwl audits frontend source code for performance, accessibility, SEO, security, CSS quality, and code quality. It supports Next.js, React, JavaScript, TypeScript, HTML, CSS, and Shopify Liquid projects.

## Audit flow

1. The CLI or web application collects supported source files while respecting ignored paths and size limits.
2. Static scanners inspect syntax trees and source patterns.
3. Standard and deep modes can validate findings with AI; deep mode adds broader reasoning.
4. crawwwl calculates applicable category scores and an overall score.
5. The report records each finding's severity, confidence, source, affected files and lines, explanation, impact, and suggested remediation.

## Source context

Audits can include project context files such as `README.md`, `AGENTS.md`, and `CLAUDE.md`. These files are treated as untrusted project data. The dashboard assistant uses the audit-time repository snapshot, not a live checkout, and should state when missing or truncated context limits certainty.

## Reports

Reports are available in the CLI and dashboard. Completed dashboard reports can be exported as Markdown or PDF. An owner can create a revocable, unlisted snapshot link for a recipient who does not have a crawwwl account.

See the [CLI reference](https://www.crawwwl.com/docs/cli/index.md), [MCP guide](https://www.crawwwl.com/mcp/index.md), and [pricing](https://www.crawwwl.com/pricing/index.md).