The Agent Skills Directory

[SAFE]: The skill provides legitimate methodologies for security researchers to analyze gRPC services. The inclusion of standard tools like grpcurl, protoc, and mitmproxy is appropriate for the stated educational and testing purpose.
[COMMAND_EXECUTION]: The skill contains various shell command examples for interacting with gRPC endpoints and analyzing binaries. These commands are intended for manual execution by a researcher and do not include patterns such as privilege escalation, persistence, or hidden data exfiltration.
[PROMPT_INJECTION]: The skill defines a surface for indirect prompt injection as it involves processing untrusted external data (such as compiled binaries and raw network messages) which could contain embedded instructions. However, this is inherent to the primary function of a security analysis skill.
Ingestion points: Processes external binaries (target_binary), JavaScript bundles (bundle.js), and protobuf messages (message.bin).
Boundary markers: No specific delimiters or warnings for embedded instructions are provided within the instructions.
Capability inventory: Uses tools like grpcurl, protoc, and python3 to analyze and interact with the data.
Sanitization: No explicit sanitization of the input binaries is described before processing by the analysis tools.

attacking-grpc-protobuf