The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on various command-line utilities for its operations, including rea, npx, jq, and system-level tools like tshark, sigrok-cli, and gdb. It explicitly provides fallback commands like npx -y rea-agents@latest for environment setup and analysis.
[INDIRECT_PROMPT_INJECTION]: The primary function of the skill is to process untrusted data from external artifacts (binaries, Android APKs, network captures) and live browser environments. Maliciously crafted content within these artifacts could potentially influence the agent's behavior via the tool outputs.
Ingestion points: External files provided by the operator and runtime data extracted from binaries, HAR/PCAP files, and browser DOM.
Boundary markers: Uses a structured 'Evidence' tracking system to maintain context, though it lacks explicit prompt-level delimiters for isolating tool output from agent instructions.
Capability inventory: Extensive capabilities including shell command execution, process instrumentation (Frida), debugging (GDB), and file system operations (extraction and bundling).
Sanitization: Includes built-in workflows for redacting secrets and credentials from browser scenario captures.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download and execution of external packages and tools. It references the rea-agents package on npm and external engines like Ghidra and JADX, which are required for full functionality.
[DATA_EXFILTRATION]: The skill is designed to handle sensitive information, such as source code, network traffic, and application metadata. It provides specific instructions for managing credentials via environment variables to prevent accidental exposure during the reverse engineering process.

reverse-engineer-anything