The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches the ysoserial-all.jar utility and clones the phpggc and JNDI-Exploit-Kit repositories from GitHub to facilitate vulnerability research.
[COMMAND_EXECUTION]: Utilizes multiple shell commands including curl, wget, java, php, python3, and git to interact with remote targets and manage local security tools.
[REMOTE_CODE_EXECUTION]: Executes the downloaded ysoserial JAR file locally to generate exploitation payloads for security testing.
[DYNAMIC_EXECUTION]: Employs Python one-liners to dynamically generate and serialize exploit objects at runtime.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from target servers (e.g., HTTP headers such as Set-Cookie and response bodies) without explicit sanitization or boundary markers, while maintaining capabilities for network access and shell execution.

hunt-deserialization