PromptGuard

Scrub sensitive data before it reaches an LLM

Detection Rules

Choose what gets replaced. The output updates in real time.

Sanitized Output

Items Redacted
0
Characters
0
Input: 0
Words
0
Input: 0
Est. Tokens
≈0
Input: ≈0

Detection is pattern-based and cannot catch everything (for example, names that are not in your keyword list). Always review the output before sharing it.