The Agent Skills Directory

[SAFE]: The skill is a specialized tool for binary triage and security auditing.
Evidence: The instructions define a clear workflow for orienting a researcher within a binary, identifying suspicious API surfaces, and pivoting through cross-references.
Evidence: The SQL recipes are tailored for detecting standard malware techniques (anti-debugging, network communication, persistence) within a target file, not for executing malicious actions on the host system.
Evidence: All external resource references, such as library names (WS2_32.dll, WININET.dll) and tools (radare2, r2xsql), are standard in the field of reverse engineering and consistent with the author 'radareorg'.
[INDIRECT_PROMPT_INJECTION]: As a binary analysis tool, the skill naturally processes untrusted data which presents a theoretical attack surface for indirect prompt injection.
Ingestion points: The tool ingests and processes metadata, strings, and symbols from potentially malicious binaries via r2xsql.
Boundary markers: There are no explicit delimiters or boundary warnings defined in the triage templates.
Capability inventory: The skill uses the Bash and Read tools to interact with the radare2 environment.
Sanitization: No specific sanitization or filtering of the binary data is described, though this is typical for low-level auditing tools where the user is looking for raw, potentially malicious content.

analysis