OpenClaw versions 2026.2.13 through 2026.3.24 contain an ANSI escape sequence injection vulnerability in approval prompts that allows attackers to spoof terminal output. Untrusted tool metadata can carry ANSI control sequences

Open WebUI has Knowledge Base Destruction and RAG Poisoning via

Langchain SQL Injection vulnerability

Langroid: SQLChatAgent _validate_query blocklist misses pg_read_file family

From versions 0.3.79 and prior and 1.0.0 to 1.0.6, a template injection vulnerability exists in LangChain's prompt template system that allows attackers to access Python object internals through template

AgenticMail gives AI agents real email addresses and phone numbers

Open WebUI: Sharing models for others to use (read permission

@dynatrace-oss/dynatrace-mcp-server has a DQL injection via parameters not

PraisonAI: Webhook signature verification skipped (fail-open) when secret unset

@dynatrace-oss/dynatrace-mcp-server has a workflow template injection via create

MCP Atlassian has SSRF via unvalidated X-Atlassian-Jira-Url

TaskWeaver has Protection Mechanism Failure and Server-Side Request Forgery

mcp-memory-service: Missing Authentication on Document API Endpoints Allows

OpenClaw: Lower-trust background runtime output is injected into trusted

npm PraisonAI AgentOS exposes unauthenticated agent listing and invocation

DeepSeek TUI has SSRF via HTTP Redirect Bypass in fetch

PraisonAI: Unauthenticated Information Disclosure of Agent Instructions via /api/agents in

Cursor is a code editor built for programming with AI

Open WebUI: Same-origin XSS to account takeover via terminal

@dynatrace-oss/dynatrace-mcp-server's create_dynatrace_notebook missing the human