[ HOME ]
>
[ BLOCKED ]
[ PRIVACY ]
#
Content was redacted by Ilvarion
If you're seeing this page, content in your AI interaction was flagged as a potential prompt injection and replaced with a redaction notice.
════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════════
##
What happened
╭
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
╮
│
Ilvarion detected patterns in content that match known prompt injection techniques. The suspicious content was replaced with:
┌
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
┐
│
[Possible prompt injection redacted by Ilvarion. See https://ilvarion.com/blocked for details]
│
└
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
┘
The rest of your content was passed through unchanged. Only the matched segments were redacted.
│
╰
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
╯
╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
##
Detection categories
Ilvarion scans for 35 patterns across these 11 categories:
[+]
Role Hijack
[+]
Instruction Override
[+]
Context Manipulation
[+]
Delimiter Attacks
[+]
Encoding Bypass
[+]
Payload Splitting
[+]
Indirect Injection
[+]
Data Exfiltration
[+]
Obfuscation
[+]
Prompt Leaking
[+]
Recursive Injection
╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
##
False positive?
Check your logs
Open the Ilvarion menubar app and check the Logs tab in Settings. Each blocked event shows the matched pattern ID and the content that triggered it.
Adjust sensitivity
You can disable specific pattern categories in the Ilvarion settings if they cause repeated false positives for your use case.
┌
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
┐
│
If you believe a pattern is incorrectly flagging legitimate content, please contact us so we can refine the detection rules.
[ Report false positive ]
│
└
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
┘
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────