What is Lyrie Research?

Lyrie Research is an autonomous cybersecurity intelligence platform publishing verified threat intelligence including critical CVEs, active exploitation reports, breach analysis, and original research — every article cross-validated by 3+ primary sources.

How does Lyrie.ai protect against rogue AI threats?

Lyrie.ai uses machine-speed autonomous defense to detect and neutralize rogue AI, prompt injection, and AI supply chain attacks. It responds before human analysts can react.

What cybersecurity topics does Lyrie Research cover?

Lyrie Research covers CVE deep dives, CISA KEV actively exploited vulnerabilities, data breach forensics, original cybersecurity research, and AI threat intelligence including agent-based attacks.

How often is Lyrie Research updated?

Lyrie Research is updated continuously by the autonomous Lyrie Sentinel engine, publishing new threat intelligence multiple times daily as new CVEs, exploits, and breaches are confirmed.

Is Lyrie Research free to access?

Yes, all articles on Lyrie Research are freely accessible. For active protection by the same intelligence engine, visit lyrie.ai.

← Home

#agent-threats

223 stories tagged.

#agent-threats

A Layered Security Framework Against Prompt Injection in RAG-Based Chatbots

AgenticMail: Unauthenticated inbound mail triggers bypassPermissions resume of the operator's Claude Code sess

PraisonAI: Server-Side Request Forgery (SSRF) in SearxNG / search_web tools via attacker-controlled searxng_ur

npm PraisonAI MCPSecurity Basic/OAuth authentication policies accept invalid credentials without validation

PraisonAI ToolsMCPServer legacy SSE transport accepts attacker Host/Origin and exposes registered tools

SafeClawBench: Separating Semantic, Audit-Evidence, and Sandbox Harm in Tool-Using LLM Agents

PhantomSkill: Malicious Code Injection in Agent Skill Ecosystems

Code-Augur: Agentic Vulnerability Detection via Specification Inference

OpenAnt: LLM-Powered Vulnerability Discovery Through Code Decomposition, Adversarial Verification, and Dynamic

Image Prompt Reconstruction Attacks on Distributed MLLM Inference Frameworks

LangChain4j: SQL injection via metadata filters in langchain4j-mariadb and langchain4j-pgvector

Claude Code: Out-of-Band Data Exfiltration via Pre-Approved HuggingFace Domain in WebFetch

OpenClaw: MCP Streamable HTTP redirects could forward configured custom headers to another origin

OTRO: Oblivious Tokenization Path with Square-Root ORAM

SoK: AI-Augmented Binary Reversing

Security and Privacy Prompts in the Wild: What Users Ask LLMs and How LLMs Respond

A Red-Team Study of Anthropic Fable 5 &amp; Opus 4.8 Models

An AI Security Agent for Banking: Multi-Vector Fraud and AML Detection Across Retail and Corporate Accounts

An Evaluation of Data Leakage Risks in Tool-Using LLM Agents in Realistic Scenarios

Seeing Is Not Screening: Multimodal Hidden Instruction Attacks on Agent Skill Scanners

Pi Agent: Potential XSS in HTML session exports via Markdown URL sanitization bypass

LangChain: Path traversal and sandbox escape in LangChain file-search middleware and loaders

Dynamic Malicious Skills in Agentic AI

Transferable Self-Evolving Playbooks for Agentic Security Auditing

How Much Can We Trust LLM Search Agents? Measuring Endorsement Vulnerability to Web Content Manipulation

SkillVetBench: LLM-as-Judge for Multi-Dimensional Security Risk Evaluation in Open-Source LLM Agent Skills

The Proxy Knows Too Much: Sealing LLM API Routers with Attested TEEs

CmdNeedle: Measuring the Incompleteness of Command Denylists for AI Agents

FragFuse: Bypassing Access Control of Large Language Model Agents via Memory-Based Query Fragmentation and Fus

From Prompts to Responses: Dual-Sided Data Leakage and Defense in Split Large Language Models

From Shield to Target: Denial-of-Service Attacks on LLM-Based Agent Guardrails

SkillMutator: Benchmarking and Defending Language-and-Code Cross-modal Attacks on LLM Agent Skills

Smarter Saboteurs, Better Fixers: Scaling &amp; Security in Linear Multi-Agent Workflows

PI-Hunter: Automated Red-Teaming for Exposing and Localizing Prompt Injections

DIG: Oracle-Guided Directed Input Generation for One-Day Vulnerabilities

SMSR: Certified Defence Against Runtime Memory Poisoning in Persistent LLM Agent Systems

MAStrike: Shapley-Guided Collusive Red-Teaming on Multi-Agent Systems

Can Open-Source LLM Agents Replace Static Application Security Testing Tools? An Empirical Assessment

Mind your key: An Empirical Study of LLM API Credential Leakage in iOS Apps

Claude Code Action: Malicious MCP Server Configuration in PRs Enables Remote Code Execution and Secret Exfiltr

Understanding and mitigating the risks of OpenClaw for non-technical users: A practical guide with Skill

Assessing Automated Prompt Injection Attacks in Agentic Environments

Securing Code Understanding: Detecting Natural Backdoor Vulnerability in Code Language Models

Game-Theoretic Multi-Agent Control for Robust Contextual Reasoning in LLMs

Training LLMs to Enforce Multi-Level Instruction Hierarchies via Gravity-Weighted Direct Preference Optimizati

Semantic Multi-Agent Intrusion Detection for IoT:Zero-Day and Adversarial Threats with Risk-Aware Reasoning

Advancing the State-of-the-Art in Empirical Privacy Auditing

MemVenom: Triggered Poisoning of Multimodal Memories in Web Agents

What the Eyes See, the LLMs Miss: Exploiting Human Perception for Adversarial Text Attacks

RAILS: Verification-Native Clearing For Agentic Commerce

Unveiling Privacy Risks in Multi-modal Large Language Models: Task-specific Vulnerabilities and Mitigation Cha

Customization under Fire: Plugin Poisoning in Text-to-Image Ecosystem

From Privacy to Workflow Integrity: Communication-Graph Metadata in Autonomous Agent Interoperability

HAVE: Host Active Verification Engine for Closing the Contextual Reality Gap in Security Digital Twins

Defending Jailbreak Attacks on Large Language Models via Manifold Trajectory Kinetics

MCP Server Kubernetes: kubectl-generic flag injection enables Kubernetes bearer token exfiltration

RedEdit: Agentic Red-Teaming of Image Safety Classifiers via MCTS-Guided Photo-Editing

SlotGCG: Exploiting the Positional Vulnerability in LLMs for Jailbreak Attacks

Steering LLM Viewpoints through Fabricated Evidence Injection

GenTI: Benchmarking LLMs for Autonomous IDPS Rule Generation for Unseen Attacks

Will the Agent Recuse Itself? Measuring LLM-Agent Compliance with In-Band Access-Deny Signals

WebMCP Tool Surface Poisoning: Runtime Manipulation Attacks on LLM Agents

Cascading Hallucination in Agentic RAG: The CHARM Framework for Detection and Mitigation

CyberGym-E2E: Scalable Real-World Benchmark for AI Agents' End-to-End Cybersecurity Capabilities

A-Live: Passive Liveness Detection via Neuromuscular Micro-Motion Signatures on Commodity Sensors

Caught in the Act(ivation): Toward Pre-Output and Multi-Turn Detection of Credential Exfiltration by LLM Agent

Description-Code Inconsistency in Real-world MCP Servers: Measurement, Detection, and Security Implications

From Control Boundary to Insurance Claim: Reconstructing AI-Mediated Losses Through the CER Framework

Bastet: A Fine-Grained Expert-Labeled Dataset for DeFi Smart Contract Vulnerability Detection

FORGE: Multi-Agent Graduated Exploitation and Detection Engineering

AI Agents Enable Adaptive Computer Worms

$π$Creds: Privately Inferred Credentials

ClawHub Security Signals: When VirusTotal, Static Analysis, and SkillSpector Disagree

Benign Inputs, Harmful Outputs: Cross-Modal Jailbreaking via Distributed Semantic Recomposition

Needles at Scale: LLM-Assisted Target Selection for Windows Vulnerability Research

Confused ChatGPT: Cross-App Context Poisoning via First-Party APIs

SS-ZKR: Spatial-Semantic Zero-Knowledge Routing for Privacy-Preserving Multi-Agent Collaboration

@agenticmail/mcp Missing Authentication for Critical Function

Automatically Attacking Software Reverse Engineering AI Agents

A Red-Team Study of Anthropic Fable 5 & Opus 4.8 Models

Smarter Saboteurs, Better Fixers: Scaling & Security in Linear Multi-Agent Workflows