The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze untrusted binary samples and memory dumps, which could contain payloads designed to exploit vulnerabilities in the analysis environment or influence the agent's logic through malformed data.
Ingestion points: Processing of external loader binaries (loader.bin), memory-dumped modules, and simulated network traffic response bytes.
Boundary markers: The skill strictly mandates the use of a sandboxed environment (re-sandbox) and network isolation (INetSim) as the primary defense mechanism.
Capability inventory: Includes file system access, process monitoring (strace, sysdig), memory dumping (gcore), and shell-based script execution.
Sanitization: Relies on environmental isolation rather than input validation.
[DYNAMIC_EXECUTION]: The skill uses Python script generation via heredoc strings (`python3
<<'EOF'`) to perform runtime validation of PE (Portable Executable) headers and section alignment for reconstructed memory modules.
[EXTERNAL_DOWNLOADS]: Fetches specialized analysis tools, specifically pe-sieve and hollows_hunter, from public GitHub releases to automate the detection and extraction of injected memory modules.

re-loader