Prompt Injection Detection

OpenRouter’s regex-based prompt injection detection scans incoming requests for common injection techniques using pattern matching. This feature is free and adds minimal latency to requests since the patterns are evaluated locally before the request is forwarded to the model provider.

How It Works

When regex-based detection is enabled on a guardrail, every incoming message is scanned against a set of patterns derived from the OWASP LLM Prompt Injection Prevention Cheat Sheet, among other resources. If a match is found, the configured action is taken:

• Flag — The request passes through unmodified; the detection is recorded for observability (metrics + analytics events) but no enforcement is applied. Useful for measuring true-positive rates on your own traffic before switching to redact or blocked.
• Redact — Matched spans are replaced with [PROMPT_INJECTION] and the sanitized request is forwarded to the model.
• Block — The entire request is rejected with a 403 before it reaches the model.

Detection Patterns

The following regex patterns are checked against all user-supplied message content. Patterns are case-insensitive unless noted otherwise.

Direct Instruction Override

Attempts to make the model ignore, override, or invalidate its original instructions.

Developer / Admin Mode Activation

Attempts to switch the model into a privileged operating mode.

System Override

Attempts to override the model's core instructions.

Prompt Extraction

Attempts to reveal the model's system prompt or internal instructions.

Role Manipulation

Attempts to change the model's identity or remove its restrictions.

DAN-Style Jailbreaks

"Do Anything Now" (DAN) and related jailbreak prompts.

Safety Bypass

Attempts to disable or circumvent the model's safety measures.

Tag Injection & Role Spoofing

Attempts to inject fake system/assistant/user tags or delimiters to confuse the model's understanding of the conversation structure.

Control Token Injection

Attempts to inject model-internal control tokens that can confuse tokenization or chat-template parsing.

Evasion Detection

In addition to the regex patterns above, the detection system includes techniques to catch common evasion strategies.

Typoglycemia Detection

Attackers may scramble the middle letters of keywords while keeping the first and last letters intact (e.g., “ignroe” instead of “ignore”). The system checks for typoglycemia variants of these target words:

Encoding-Based Evasion

The system decodes Base64 and hex-encoded content (including space-separated hex pairs like 69 67 6e 6f 72 65), then checks the decoded text for injection keywords:

This catches attempts to hide malicious instructions behind encoding layers. Two encoding detectors run: base64_encoded_injection, hex_encoded_injection.

Character-Spaced Evasion

Text with character spacing (e.g., i g n o r e p r e v i o u s) is normalized by collapsing spaces, then re-scanned against all patterns. This prevents simple spacing-based evasion.

Limitations

• Regex-based detection is not exhaustive. Sophisticated or novel injection techniques may not be caught.
• Flag mode does not enforce. A flagged request is forwarded to the model exactly as submitted — the detection is recorded for dashboards and analytics only. Use flag to measure match rates on real traffic; switch to redact or block once you’re confident the false-positive rate is acceptable.
• False positives are possible. Some legitimate prompts may contain phrases that match these patterns (e.g., a prompt about security testing). Test your guardrail configuration with representative traffic — ideally in flag mode first — before enforcing broadly.

Further Reading

• OWASP LLM Prompt Injection Prevention Cheat Sheet
• Guardrails documentation
• Guardrails API reference