CI and pull requests
Run a scan after preview deployment, then fail or flag the build when critical launch blockers return.
PageLens AI CLI - npm package
Run PageLens AI audits from your terminal, CI, and deploy flow.
The PageLens AI CLI turns launch QA into a repeatable command. Use it for preview deployments, client handoff checks, agency retainers, and release gates where opening the dashboard manually slows you down. Add --markdown when you want the same agent-ready fix queue the result exports for Codex, Cursor, Claude Code, Copilot, Windsurf, Lovable, Bolt, Replit or v0. CLI and API automation are included on Builder, Pro and Agency plans.
Requires a PageLens AI API key for account scans. API keys require Builder or higher and live in Settings - Integrations.
Where the CLI fits.
The dashboard is where you review and share reports. The CLI is for the moments where PageLens AI should run automatically. Pin the GitHub Action to a release tag such as v1rather than a moving branch when you use it in production. Use --builder and --moment when you want the resulting report and fix prompts framed for the AI tool and launch risk you are actually working through. Use --markdown pagelens-report.md to save a ready-to-paste priority fix pack for that tool after the scan completes.
Deploy checks
Pair the CLI with Vercel, GitHub Actions, or your release script so PageLens AI becomes a repeatable gate.
Recurring client QA
Agencies can run the same command for every client site before handoff, monthly review, or campaign launch.
Built for agency and client work.
Agencies do not need another one-off dashboard ritual. They need a repeatable check that can run across client sites before launch, before campaign spend, and before monthly reporting.
Run the same audit command for every client handoff
Keep proof that launch checks happened before sign-off
Use reports in client review meetings, not just engineering
Trigger checks from deploy previews and release scripts
CLI, API, deploy hooks, or MCP?
They are complementary surfaces for the same PageLens AI data.
Browser UI
Best when you are exploring a report, sharing findings, or choosing scan options manually.
CLI
Best when the scan should happen automatically from a terminal, CI job, deploy preview, or agency checklist, with optional Markdown output for your coding agent.
MCP
Best when Cursor, Claude Code, Codex, or another agent needs to read PageLens AI findings and work the fix list with context.