Articles
Daily digests summarizing the vulnerabilities and incidents tracked by VulnWatch.
-
· 45 entries · daily · Subscribers only
VulnWatch Daily: Critical RCE in SGLang, Ollama, and PraisonAI Agent Ecosystem
October 8, 2026: A wave of critical vulnerabilities exposes AI infrastructure to remote code execution. SGLang, Ollama, and the PraisonAI framework face severe flaws allowing unauthenticated access, secret exfiltration, and full system compromise via prompt injection and deserialization bugs.
-
· 40 entries · daily · Subscribers only
VulnWatch Daily: Langflow RCE Flood, MCP Protocol Flaws, and Agent Injection Risks
A critical surge in Langflow vulnerabilities exposes AI orchestration platforms to remote code execution. New findings also highlight severe risks in Model Context Protocol implementations and agent webhook authentication.
-
· 17 entries · daily · Subscribers only
VulnWatch Daily: Critical RCE in Langflow & MCP OAuth Flaws
Today's digest highlights critical remote code execution risks in Langflow and Progress Software, alongside severe OAuth credential leakage in the Model Context Protocol SDK. Immediate patching is required for authenticated RCE and SSRF vectors.
-
· 44 entries · daily · Subscribers only
VulnWatch Daily: Langflow RCE Storm, vLLM DoS Vectors, and Docling Supply Chain Risks
A catastrophic cluster of vulnerabilities in Langflow enables trivial RCE via MCP and webhook bypasses. Simultaneously, vLLM faces multiple DoS vectors, and Docling exhibits supply chain and template injection flaws requiring immediate attention.
-
· 63 entries · weekly
VulnWatch Weekly: Agentic RCE Epidemic & The Fall of Vibe-Trading
This week's digest highlights a catastrophic cluster of Remote Code Execution (RCE) vulnerabilities in agentic frameworks, led by critical flaws in Vibe-Trading and Devika. Security teams must immediately audit LLM tool definitions and sandbox boundaries as the 'agent-as-a-service' model faces its first major exploitation wave.
-
· 2 entries · daily · Subscribers only
VulnWatch Daily: SSRF in SciPhi R2R and Info Leak in Weaviate Verba
Today's digest covers two medium-severity vulnerabilities affecting AI infrastructure: an SSRF in SciPhi-AI R2R allowing remote API manipulation and an information disclosure flaw in Weaviate Verba exposing environment variables.
-
· 3 entries · daily · Subscribers only
VulnWatch Digest: RCE in OpenAM, DoS in SGLang, and SEO Poisoning
Today's digest covers critical risks in AI infrastructure: an unauthenticated RCE vector in OpenAM's legacy SOAP interface, a handshake DoS in Mooncake/SGLang, and an authorization bypass in WordPress AI plugins enabling content poisoning.
-
· 12 entries · daily · Subscribers only
VulnWatch Daily: Critical Agentic RCE Chains and MCP SSRF Flaws
Today's digest highlights critical RCE chains in Vibe-Trading agents, WebSocket hijacking in Headroom, and emerging SSRF risks in Model Context Protocol implementations.
-
· 19 entries · daily · Subscribers only
VulnWatch Daily: Agentic MCP Flaws, LiteLLM Auth Gaps, and vLLM DoS
Today's digest highlights critical authorization bypasses in Model Context Protocol (MCP) servers, a cascade of authentication flaws in LiteLLM proxies, and new denial-of-service vectors in vLLM audio processing.
-
· 8 entries · daily · Subscribers only
VulnWatch Daily: Critical Sandbox Escapes and Agentic Auth Flaws
Today's digest highlights critical sandbox escapes in Kobako, severe cross-tenant auth flaws in Soft Machine, and RCE risks in MCP tools. Immediate patching is required for agentic platforms and LLM runtimes.
-
· 7 entries · daily · Subscribers only
VulnWatch Digest: Critical Prompt Injections & AI-Generated Code Flaws
Today's digest highlights critical prompt injection risks in Meta's GenAI stack and Ollama, alongside severe memory corruption in embedded AI systems traced back to AI-generated code.
-
· 11 entries · daily · Subscribers only
VulnWatch Daily: Agentic RCE Surge in MCP Servers and LiteLLM Auth Flaws
Today's digest highlights critical command injection flaws in Model Context Protocol (MCP) servers, a severe authentication bypass in LiteLLM, and multiple sandbox escape risks in DeepSeek harnesses requiring immediate platform review.
-
· 120 entries · weekly
VulnWatch Weekly: The MCP Authentication Crisis & Decepticon's Role Forgery
This week's digest highlights a catastrophic wave of authentication bypasses in Model Context Protocol (MCP) servers, led by critical flaws in mcp-atlassian. We also analyze 'Decepticon,' a new prompt injection vector enabling role-boundary forgery via ChatML literals.
-
· 8 entries · daily · Subscribers only
Agentic Supply Chain Crisis: Critical MCP Vulnerabilities in Obot, Penpot, and Heym
A surge of critical vulnerabilities in Model Context Protocol (MCP) implementations exposes AI agents to unauthorized access, SSRF, and data leakage. Immediate patching required for Obot, Penpot, and Heym platforms.
-
· 16 entries · daily · Subscribers only
VulnWatch Digest: Critical MCP Flaws and vLLM DoS Risks
Today's digest highlights severe authorization bypasses in Model Context Protocol (MCP) servers and a cluster of denial-of-service vulnerabilities in vLLM affecting multimodal inference stability.