---
title: "Joey Melo (@pangea-joeymelo) | Hashnode"
description: "Joey Melo (@pangea-joeymelo) writes on Hashnode. Read their latest articles and follow their work."
url: "https://hashnode.com/@pangea-joeymelo"
---

@pangea-joeymelo

# Joey Melo

@pangea-joeymelo · Joined July 2025

### [Decoding LLM Attack Surfaces: A Deep Dive into Model Vulnerabilities](https://pangea.cloud/blog/decoding-llm-attack-surfaces-a-deep-dive-into-model-vulnerabilities?utm_source=hashnode&utm_medium=feed)

The attack surface of Large Language Models (LLMs) is a critical area of research, especially as these models become more integrated into various applications. To shed light on this, I conducted a series of tests on four prominent models: GPT-5, Gemi...

4 min read · [pangea.cloud](https://pangea.cloud/blog/decoding-llm-attack-surfaces-a-deep-dive-into-model-vulnerabilities?utm_source=hashnode&utm_medium=feed)

[Aug 28, 2025](https://hashnode.com/posts/decoding-llm-attack-surfaces-a-deep-dive-into-model-vulnerabilities/68b09b2cbb06550fceec596e)

### [Mining the Index: Uncovering Sensitive Data in Public ChatGPT Histories via Google Search](https://pangea.cloud/blog/mining-the-index-uncovering-sensitive-data-in-public-chatgpt-histories-via-google-search?utm_source=hashnode&utm_medium=feed)

Learn how ChatGPT shared conversation histories are being indexed by Google, exposing sensitive data, and the implications for privacy and security

3 min read · [pangea.cloud](https://pangea.cloud/blog/mining-the-index-uncovering-sensitive-data-in-public-chatgpt-histories-via-google-search?utm_source=hashnode&utm_medium=feed)

[Aug 1, 2025](https://hashnode.com/posts/mining-the-index-uncovering-sensitive-data-in-public-chatgpt-histories-via-google-search/688cf2f2c48928c34ab1d3b0)

### [Phishing for Gemini: Invisible Instructions to Manipulate Email Summaries](https://pangea.cloud/blog/phishing-for-gemini-invisible-instructions-to-manipulate-email-summaries?utm_source=hashnode&utm_medium=feed)

Research uncovers AI phishing attacks, revealing AI security vulnerabilities via covert manipulation and hidden prompts

3 min read · [pangea.cloud](https://pangea.cloud/blog/phishing-for-gemini-invisible-instructions-to-manipulate-email-summaries?utm_source=hashnode&utm_medium=feed)

[Jul 17, 2025](https://hashnode.com/posts/phishing-for-gemini-invisible-instructions-to-manipulate-email-summaries/68791adec6447367308a71e0)

### [Bypassing AI Malware Analysis with Prompt Injection](https://pangea.cloud/blog/bypassing-ai-malware-analysis-with-prompt-injection?utm_source=hashnode&utm_medium=feed)

Discover how prompt injection can bypass AI malware detection, revealing new vulnerabilities in tools like gemini-cli and GitHub Copilot

2 min read · [pangea.cloud](https://pangea.cloud/blog/bypassing-ai-malware-analysis-with-prompt-injection?utm_source=hashnode&utm_medium=feed)

[Jul 11, 2025](https://hashnode.com/posts/bypassing-ai-malware-analysis-with-prompt-injection/68714b019194458bab99c794)

### [The Hidden Influence: Prompt Injection in ArXiv Research Papers](https://pangea.cloud/blog/the-hidden-influence-prompt-injection-in-arxiv-research-papers?utm_source=hashnode&utm_medium=feed)

Discover how prompt injections in ArXiv research papers manipulate AI reviews and the potential security risks they pose

2 min read · [pangea.cloud](https://pangea.cloud/blog/the-hidden-influence-prompt-injection-in-arxiv-research-papers?utm_source=hashnode&utm_medium=feed)

[Jul 11, 2025](https://hashnode.com/posts/the-hidden-influence-prompt-injection-in-arxiv-research-papers/68714a2f67f7c5c21defe4a4)
