arlong
Sign in

YOUR WORKSPACE / ARLONG SCREEN

Inspect before you ingest.

Boundary screens the original input, isolates hostile content, and checks what reaches your agent.

API reference ↗
INPUT / 1 CREDIT PER CHECK

Content to screen

Examples are editable and free to load. Each submitted check uses the same engine and account balance as the API.

0 charactersNever paste credentials or private account keys into sample content.
Request options

The URL limit bounds extracted evidence. Text screening inspects up to 250,000 characters; review sanitization metadata for limits and removals.

The key stays in this page's memory and is sent only to Arlong. Or sign in to use your account directly. Get a key ↗

Original and retained content are screened separately. Only allowed, cleaned evidence is released.

BOUNDARY / DELIVERY REPORT

What your agent gets

[ ]

No check yet

Load an example or paste content, then run Screen. Reports appear here without executing or rendering submitted HTML.

Use screening as one layer.

A decision is not a guarantee against unseen attacks. Keep tool permissions scoped and authorize sensitive actions separately. Integration checklist ↗