deserialization-php

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly instructs embedding secrets verbatim (e.g., using APP_KEY in encryption commands like laravel_crypto_killer.py -k "base64:APP_KEY_HERE") and requires reporting "new credentials or tokens found" in the return summary, which forces the LLM to handle/output secret values directly.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). These URLs show direct phar:// references to uploaded image files (classic PHAR polyglot deserialization exploitation) alongside a link to phpggc (a gadget-chain generator used to build RCE payloads), indicating an explicit exploit workflow and therefore a high-risk/suspicious download/execution source.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This document is an explicit, actionable exploitation guide for PHP deserialization that details generating gadget chains (phpggc), PHAR polyglots, cookie forgery, and techniques to achieve remote code execution, arbitrary file writes and persistence—clearly enabling backdoor/RCE abuse.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs cloning and using phpggc from https://github.com/ambionics/phpggc (git clone ...), and then runs phpggc commands as part of exploitation, so remote code would be fetched and executed as a required runtime dependency.