Skip to content

snailsploit[$]Adversarial · Research

prompt injection · hub

7 writeups · updated 2026.05

Prompt Injection.

Direct injection, indirect injection through retrieved content, MCP tool poisoning, memory-instruction injection across sessions, and the trust-boundary failures that make each one work. Pattern reference plus original research on novel chains.

Semantic-to-Metadata Smuggling in Multi-LoRA Routing Gateways

Signal suppression at the gateway: attacker-controlled paraphrase selects which adapter — and which safety policy — serves the next token. SHELL.003.

prompt-injection

Indirect Injection Was Never Blind

The SSE stream carries the tool call before the interface hides it — closed-loop indirect injection via render-layer redaction. SHELL.002.

prompt-injection

Memory Injection Through Nested Skills

Skill injection plus cross-session memory poisoning creates a self-healing LLM implant. Explore this novel persistence chain exploiting agent trust boundaries.

prompt-injection

Prompt Injection Examples

Real-world prompt injection examples across direct injection, indirect injection, MCP tool poisoning, and memory attacks. Learn how each pattern works.

prompt-injection

MCP Security Hardening

How to secure MCP servers in production AI environments. Real-world vulnerability scenarios, server configuration hardening, and AATMF-mapped defense patterns.

prompt-injection

Custom Instruction Backdoor

Uncovering emergent prompt injection risks through ChatGPT custom instructions. Learn how user settings become persistent attack vectors in LLM applications.

prompt-injection

MCP Threat Analysis

Offensive threat analysis of the Model Context Protocol. Attack chains, privilege escalation paths, and exploitation techniques targeting AI tool integration.

prompt-injection

More research

all research →wiki →cves →