exploitation-knowledge

SUSPICIOUS/HIGH RISK. The skill is purpose-built to help an AI agent perform offensive exploitation against remote targets, including RCE, shell acquisition, brute force, and post-compromise file access. Its raw-GitHub PoC download guidance adds supply-chain risk, but the main issue is that the capability itself is an AI-agent exploitation toolkit, not a narrowly scoped benign integration.

This Markdown skill/playbook contains highly actionable offensive exploitation instructions rather than defensive or safe educational material. It includes multiple reverse-shell payload variants (bash/Python/PHP/netcat/named pipes), web-shell command-execution workflows, credential abuse tooling guidance, explicit sensitive file/flag/data capture steps (e.g., user.txt and /etc/passwd), and evasion guidance (e.g., Base64→bash and alternative execution paths). While it is not a stealthy self-contained malware binary, the distributed content meaningfully enables unauthorized compromise and data theft, making the security risk extremely high for any general-use package context.