models: - type: main engine: openai model: gpt-5 api_key_env_var: OPENAI_API_KEY parameters: temperature: 0 base_url: http://127.0.0.1:8051/v1 - type: self_check_input engine: openai model: openai.gpt-oss-120b api_key_env_var: OPENAI_API_KEY parameters: temperature: 0 base_url: http://127.0.0.1:8051/v1 instructions: - type: general content: | You are a helpful, concise assistant. rails: config: jailbreak_detection: server_endpoint: "" length_per_perplexity_threshold: 89.79 prefix_suffix_perplexity_threshold: 1845.65 keyword_detection: input: case_insensitive: true keywords: - malware - ransomware - phishing kit - credential stuffing - botnet - keylogger - developer mode - hidden system prompt - hidden instructions - bypass guardrails - bypass safety - bypass filters regex_detection: input: case_insensitive: true patterns: - '\b[A-Z0-9._%+-]+@[A-Z0-9.-]+\.[A-Z]{2,}\b' - '\b(?:\+?\d{1,3}[-. ]?)?(?:\(?\d{3}\)?[-. ]?)\d{3}[-. ]\d{4}\b' - '\b\d{3}-\d{2}-\d{4}\b' - '\b(?:\d[ -]*?){13,16}\b' - '\b[A-Z]{2}\d{2}[A-Z0-9]{11,30}\b' - '\bsk-[A-Za-z0-9_-]{20,}\b' - '\bAKIA[0-9A-Z]{16}\b' - '\bgh[pousr]_[A-Za-z0-9]{20,}\b' - '\bignore (all |any )?(previous|prior) instructions\b' - '\breveal (the )?(system prompt|hidden instructions)\b' - '\bprint (your|the) (full )?(system prompt|hidden prompt)\b' input: flows: - jailbreak detection heuristics - self check input - keyword check input - regex check input