TokenPad
Safety

Prompt Injection Scanner

Check untrusted text for known injection patterns before it enters a prompt.

Your input

Paste untrusted text: a retrieved document, a user upload, a tool result.

189 characters3 lines0 tokensor drop a file

Injection ScannerExact
0Patterns matched
Patterns matched0no known patterns
High severity0override or exfiltration attempts
Rules checked10pattern matching, not detection
Token cost of this result
Output tokens0
As input$0.00
× 100K requests$0.00

Everything on this page runs in your browser. Nothing you paste is transmitted, because there is no server here to transmit it to.

Result
 

What prompt injection is

Any text that reaches a model is a candidate instruction. If your application inserts a retrieved document, a user upload or a scraped page into a prompt, whoever wrote that content can attempt to address the model directly.

The consequences scale with what the model can do. For a summariser it is embarrassing. For an agent with tools it is a security incident, because the injected instruction can drive real actions.

What this scanner can and cannot do

It matches published patterns: the phrasings that appear repeatedly in documented attacks. That covers opportunistic and copy-pasted attempts, which is the majority of what most applications encounter.

It will not catch a novel or paraphrased injection, and it cannot catch one written in a language it does not pattern-match. A clean result is not a safety guarantee — treat untrusted text as untrusted regardless of what this says.

What actually reduces the risk

Structural defences beat detection. Keep untrusted content inside a delimited block and state plainly that content within it is data, not instructions. Never let retrieved text sit adjacent to your own instructions without a boundary.

Then limit the blast radius: give agents the narrowest tool permissions that work, require confirmation for anything irreversible, and never place secrets in a context that untrusted content also occupies.

Detection is a useful additional layer. It is not the layer that saves you.

Frequently asked questions

Can prompt injection be solved completely?
Not currently. Instructions and data share one channel, and no reliable general defence exists. The practical approach is layered: delimit, instruct, limit permissions, scan, and assume some attempts will succeed.
Does this catch indirect injection?
It catches the patterns wherever they appear, including in retrieved documents — which is where indirect injection lives. Scan retrieved content, not just what the user typed.
What are the highest-risk applications?
Agents with tool access, especially anything that sends email, writes to a database, or spends money. There, an injection converts from a content problem into an action problem.

More safety tools