5018 matches found
Microsoft Azure DevOps MCP Flaw Lets Hidden PR Comments Hijack AI Review Agents
A single invisible comment in an Azure DevOps pull request can turn a reviewer's own AI coding agent against them, driving it into projects the attacker has no rights to reach and quietly leaking what it finds. The flaw is in Microsoft's official Azure DevOps MCP server, and it works because one ...
SecureAI-Scan v0.4.1
SecureAI-Scan The AI security scanner that proves its findings. SecureAI-Scan finds LLM, MCP, Agent Skill, and RAG vulnerabilities in TypeScript, JavaScript, and Python — and shows you the evidence: the exact source → flow → sink path for every dataflow finding, resolved through real imports, not...
PT-2026-63365
A flaw was found in the Ansible Lightspeed Model Context Protocol MCP server. This vulnerability, known as path traversal, allows an attacker to manipulate an AI agent through indirect prompt injection. By doing so, the attacker can cause the server to write files to unauthorized locations on the...
CVE-2026-44192: Improper Limitation of a Pathname to a Restricted Directory
A flaw was found in the Ansible Lightspeed Model Context Protocol MCP server. This vulnerability, known as path traversal, allows an attacker to manipulate an AI agent through indirect prompt injection. By doing so, the attacker can cause the server to write files to unauthorized locations on the...
EUVD-2026-46421
mcp-webresearch 0.1.7 contains a server-side request forgery vulnerability that allows attackers to access internal network services by supplying loopback, link-local, or cloud metadata addresses to the visitpage tool, which only validates the URL protocol without filtering private or reserved IP...
CVE-2026-65056
mcp-webresearch 0.1.7 contains a server-side request forgery vulnerability that allows attackers to access internal network services by supplying loopback, link-local, or cloud metadata addresses to the visitpage tool, which only validates the URL protocol without filtering private or reserved IP...
CVE-2026-65056
CVE-2026-65056 describes a server-side request forgery in mcp-webresearch 0.1.7. The vulnerability arises because visit_page validates only URL protocol and does not filter private or reserved IP ranges, enabling an attacker to access internal network services by supplying loopback, link-local, o...
CVE-2026-65056 mcp-webresearch Server-Side Request Forgery in visit_page Due to Missing Internal-IP Filtering
mcp-webresearch 0.1.7 contains a server-side request forgery vulnerability that allows attackers to access internal network services by supplying loopback, link-local, or cloud metadata addresses to the visitpage tool, which only validates the URL protocol without filtering private or reserved IP...
CVE-2026-65056 mcp-webresearch Server-Side Request Forgery in visit_page Due to Missing Internal-IP Filtering
mcp-webresearch 0.1.7 contains a server-side request forgery vulnerability that allows attackers to access internal network services by supplying loopback, link-local, or cloud metadata addresses to the visitpage tool, which only validates the URL protocol without filtering private or reserved IP...
CVE-2026-47395
PraisonAI is a multi-agent teams system. Prior to version 4.6.40 of PraisonAI, corresponding to version 1.6.40 of praisonaiagents, PraisonAI's direct-prompt CLI automatically expands @url: mentions in raw prompt text before agent execution begins. If a prompt contains @url:, the CLI calls...
CVE-2026-47395 PraisonAI CLI automatically resolves @url mentions in prompt text and can read loopback URLs into model context
PraisonAI is a multi-agent teams system. Prior to version 4.6.40 of PraisonAI, corresponding to version 1.6.40 of praisonaiagents, PraisonAI's direct-prompt CLI automatically expands @url: mentions in raw prompt text before agent execution begins. If a prompt contains @url:, the CLI calls...
EUVD-2026-46303
PraisonAI is a multi-agent teams system. Prior to version 4.6.40 of PraisonAI, corresponding to version 1.6.40 of praisonaiagents, PraisonAI's direct-prompt CLI automatically expands @url: mentions in raw prompt text before agent execution begins. If a prompt contains @url:, the CLI calls...
CVE-2026-47395
CVE-2026-47395 – PraisonAI CLI loopback/SSRF via @url: mentions Prior to PraisonAI v4.6.40 (and praisonaiagents v1.6.40), the direct-prompt CLI automatically expands @url:(http|https) prompts, performing an unrestricted urllib.request.urlopen() to the attacker-controlled URL. The fetched response...
New ClickLock Stealer locks your Mac until you hand over your password
ClickLock Stealer is a new, modular macOS infostealer delivered via ClickFix-style phishing pages that can lock a victim’s Mac, steal their macOS password, browser and password manager data, cryptocurrency wallets, and then leave behind a persistent backdoor. The malware was discovered by Group-I...
Twin Agent: Context Residual Compression for Privilege Separated Agents
Large language model LLM agents are vulnerable to security risks, such as prompt injection attacks from untrusted context that manipulate downstream reasoning and tool use. Existing secure-by-design approaches mitigate this risk by separating untrusted observations from privileged execution and...
CVE-2026-65056: Server-Side Request Forgery (SSRF)
mcp-webresearch 0.1.7 contains a server-side request forgery vulnerability that allows attackers to access internal network services by supplying loopback, link-local, or cloud metadata addresses to the visitpage tool, which only validates the URL protocol without filtering private or reserved IP...
CVE-2026-47395: Exposure of Sensitive Information to an Unauthorized Actor
PraisonAI is a multi-agent teams system. Prior to version 4.6.40 of PraisonAI, corresponding to version 1.6.40 of praisonaiagents, PraisonAI's direct-prompt CLI automatically expands @url: mentions in raw prompt text before agent execution begins. If a prompt contains @url: , the CLI calls...
PT-2026-62118
mcp-webresearch 0.1.7 contains a server-side request forgery vulnerability that allows attackers to access internal network services by supplying loopback, link-local, or cloud metadata addresses to the visit page tool, which only validates the URL protocol without filtering private or reserved I...
CVE-2026-57495
AgenticMail gives AI agents real email addresses and phone numbers. In @agenticmail/claudecode prior to version 0.2.39, @agenticmail/codex prior to version 0.1.33, @agenticmail/core prior to version 0.9.43, and @agenticmail/openclaw prior to version 0.5.71, two inbound-mail handlers act on a...