nintechio/aitdp

Adapters: Rebuff, NeMo Guardrails, Prompt Guard (HF) about 2 hours ago

help wanted adapter

0

AI Threat Detection Protocol — open spec, community ruleset and reference SDK for detecting prompt injection, jailbreaks, data exfiltration and tool abuse in LLM apps and AI agents.

Python

#agentic-ai#ai-agents#ai-security#cli#detection-rules#guardrails#jailbreak-detection#langchain#llm#llm-security#mcp#mitre-atlas#openai#owasp-llm-top-10#prompt-injection#protocol#python#sdk#security-tools#threat-detection

TypeScript SDK (@aitdp/core) with bundled rules about 2 hours ago

enhancement help wanted typescript

nintechio/aitdp

0

AI Threat Detection Protocol — open spec, community ruleset and reference SDK for detecting prompt injection, jailbreaks, data exfiltration and tool abuse in LLM apps and AI agents.

Python

#agentic-ai#ai-agents#ai-security#cli#detection-rules#guardrails#jailbreak-detection#langchain#llm#llm-security#mcp#mitre-atlas#openai#owasp-llm-top-10#prompt-injection#protocol#python#sdk#security-tools#threat-detection

Benchmark: ingest public prompt-injection datasets about 2 hours ago

enhancement help wanted

nintechio/aitdp

0

AI Threat Detection Protocol — open spec, community ruleset and reference SDK for detecting prompt injection, jailbreaks, data exfiltration and tool abuse in LLM apps and AI agents.

Python

#agentic-ai#ai-agents#ai-security#cli#detection-rules#guardrails#jailbreak-detection#langchain#llm#llm-security#mcp#mitre-atlas#openai#owasp-llm-top-10#prompt-injection#protocol#python#sdk#security-tools#threat-detection

Rule: PII leakage (emails, phone numbers, national IDs, IBANs) in model output about 2 hours ago

good first issue rule

nintechio/aitdp

0

AI Threat Detection Protocol — open spec, community ruleset and reference SDK for detecting prompt injection, jailbreaks, data exfiltration and tool abuse in LLM apps and AI agents.

Python

#agentic-ai#ai-agents#ai-security#cli#detection-rules#guardrails#jailbreak-detection#langchain#llm#llm-security#mcp#mitre-atlas#openai#owasp-llm-top-10#prompt-injection#protocol#python#sdk#security-tools#threat-detection

Rule: multi-agent privilege escalation / agent-to-agent injection about 2 hours ago

help wanted rule

nintechio/aitdp

0

AI Threat Detection Protocol — open spec, community ruleset and reference SDK for detecting prompt injection, jailbreaks, data exfiltration and tool abuse in LLM apps and AI agents.

Python

#agentic-ai#ai-agents#ai-security#cli#detection-rules#guardrails#jailbreak-detection#langchain#llm#llm-security#mcp#mitre-atlas#openai#owasp-llm-top-10#prompt-injection#protocol#python#sdk#security-tools#threat-detection

Rule: RAG poisoning markers in retrieved documents about 2 hours ago

help wanted rule

nintechio/aitdp

0

AI Threat Detection Protocol — open spec, community ruleset and reference SDK for detecting prompt injection, jailbreaks, data exfiltration and tool abuse in LLM apps and AI agents.

Python

#agentic-ai#ai-agents#ai-security#cli#detection-rules#guardrails#jailbreak-detection#langchain#llm#llm-security#mcp#mitre-atlas#openai#owasp-llm-top-10#prompt-injection#protocol#python#sdk#security-tools#threat-detection

Rule: indirect injection via HTML comments and hidden elements about 2 hours ago

good first issue rule

nintechio/aitdp

0

AI Threat Detection Protocol — open spec, community ruleset and reference SDK for detecting prompt injection, jailbreaks, data exfiltration and tool abuse in LLM apps and AI agents.

Python

#agentic-ai#ai-agents#ai-security#cli#detection-rules#guardrails#jailbreak-detection#langchain#llm#llm-security#mcp#mitre-atlas#openai#owasp-llm-top-10#prompt-injection#protocol#python#sdk#security-tools#threat-detection

Rule: fake role/turn delimiters ("### System:", "<|im_start|>system", "[INST]") about 2 hours ago

good first issue rule

nintechio/aitdp

0

AI Threat Detection Protocol — open spec, community ruleset and reference SDK for detecting prompt injection, jailbreaks, data exfiltration and tool abuse in LLM apps and AI agents.

Python

#agentic-ai#ai-agents#ai-security#cli#detection-rules#guardrails#jailbreak-detection#langchain#llm#llm-security#mcp#mitre-atlas#openai#owasp-llm-top-10#prompt-injection#protocol#python#sdk#security-tools#threat-detection

Rule: multilingual instruction-override injection (es, de, fr, pt, it) about 2 hours ago

good first issue rule

nintechio/aitdp

0

AI Threat Detection Protocol — open spec, community ruleset and reference SDK for detecting prompt injection, jailbreaks, data exfiltration and tool abuse in LLM apps and AI agents.

Python

#agentic-ai#ai-agents#ai-security#cli#detection-rules#guardrails#jailbreak-detection#langchain#llm#llm-security#mcp#mitre-atlas#openai#owasp-llm-top-10#prompt-injection#protocol#python#sdk#security-tools#threat-detection

Open Issues Need Help