The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to install standard security tools from official registries including PyPI (pwntools, ropper, ROPgadget, angr) and RubyGems (seccomp-tools).
[COMMAND_EXECUTION]: Utilizes system package managers and security-focused CLI tools for binary analysis, gadget searching, and environment setup.
[PRIVILEGE_ESCALATION]: Recommends the use of sudo with system package managers for the purpose of installing required tools and dependencies.
[DYNAMIC_EXECUTION]: Involves the creation and execution of Python-based exploit scripts using the pwntools framework to verify vulnerabilities against target processes.
[INDIRECT_PROMPT_INJECTION]: The skill processes external binary files as input, presenting an attack surface for indirect prompt injection. This is addressed by explicit guidance to perform dynamic verification within a sandbox environment.
Ingestion points: Target binaries specified for analysis (e.g., ./target).
Boundary markers: Explicitly references a sandbox environment for dynamic validation.
Capability inventory: Includes powerful analysis and script execution frameworks like pwntools, ropper, and angr.
Sanitization: Relies on sandboxed execution and manual oversight by the researcher during the exploit development process.

re-exploit