The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the ghidra-agent-cli tool along with standard shell commands such as mktemp, cp, and cat to manage analysis data and update project artifacts within the artifacts/ and analysis/ directories.
[SAFE]: The skill instructions enforce a strict, logical order of operations (types, constants, vtables, identity, then decompilation), which is a best practice in reverse engineering to ensure accuracy and reduce hallucinations based on raw decompilation.
[SAFE]: No evidence of obfuscation, multi-layer encoding, or hidden content was found in the instructions or metadata across all language versions.
[SAFE]: All network operations and tool invocations are performed via the ghidra-agent-cli tool, which is consistent with the vendor's provided infrastructure (ByteLandTechnology).
[SAFE]: The skill does not attempt to access sensitive system locations, user credentials, or perform unauthorized remote code execution.

headless-ghidra-analyze-function