prompt-injection
Safeguard articles tagged "prompt-injection" — guides, analysis, and best practices for software supply chain and application security.
135 articles
MCP-Led Automation: Securing Model Context Protocol Workflows
MCP-led agent workflows hand real tools to a language model. That power is also the attack surface. Here is how tool poisoning works and how to defend against it.
EchoLeak (CVE-2025-32711): The First Zero-Click LLM Exfiltration in Production
Aim Security's CVE-2025-32711 exfiltrated Microsoft 365 Copilot data via a single crafted email. The XPIA classifier failed, CSP let attackers through, and CVSS 9.3 followed.
EchoLeak (CVE-2025-32711): The First Zero-Click Production LLM Exfiltration
A single crafted email could exfiltrate data from Microsoft 365 Copilot without a user click. We walk the attack chain, the patch, and the lessons for agent operators.
Prompt-injection vectors specific to MCP servers and how to layer defenses
MCP servers expose three distinct prompt-injection surfaces — resource contents, tool outputs, and sampling requests — and each one needs its own defense layer. Here is how to think about them together.
MCP tool poisoning: hidden instructions and rug-pulled tool definitions
Tool-poisoning attacks against Model Context Protocol servers hide adversarial instructions inside tool descriptions and silently mutate tool definitions after install. Here is how the attack works and how to defend against it.
Prompt injection attacks against AI coding/security tools
AI coding assistants like Copilot and Cursor can be hijacked by hidden text in files, comments, and packages. Here's how prompt injection malware works and how Safeguard detects it.
What Is Prompt Engineering? A Security Guide for LLM Applications
Prompt engineering is how you steer an LLM, and it is also where a lot of application security now lives. Here is how to write prompts that resist injection and leakage.
CISA's Agentic AI Secure Adoption Guide (May 2026): What It Means for Software Supply Chains
On May 4, 2026, CISA and international partners published guidance on the secure adoption of agentic AI. We break down the named risks, the recommended controls, and how to operationalize them for AppSec and platform teams.
GitHub MCP Server Private-Repo Exfiltration: The May 2025 Invariant Labs Disclosure
Invariant Labs showed that a malicious GitHub Issue could hijack any MCP-connected agent into leaking private-repo contents. The architecture, not a bug, is the problem.
Jailbreak Meaning: What It Is in AI and Devices
Jailbreak has two meanings today: removing restrictions on a device, and tricking an AI model into ignoring its safety rules. This guide covers both.
What Is AI Jailbreaking? A Defender's Security Guide
To jailbreak AI means to bypass a model's safety guardrails with crafted prompts. Here is how the technique works and, more usefully, how to defend against it.
Line Jumping: How MCP Tool Descriptions Attack Before Tools Are Called
Trail of Bits coined 'line jumping' for prompt injection delivered through MCP tool descriptions on connection. It bypasses every tool-invocation guardrail by design.
Self-healing security runs on Safeguard.
Your first fix PR is minutes away.
No sales call required, even your agent can complete the purchase over MCP.