The Agent Skills Directory

[SAFE]: The skill documents the use of the ida_domain Python library for reverse engineering tasks. All external references point to hex-rays.com, the official domain for the tool's vendor.
[DATA_EXFILTRATION]: No evidence of unauthorized data transmission was found. The skill includes a security-conscious recommendation to disable Lumina (a remote metadata service) by redirecting its host to 0.0.0.0 to prevent unintended network lookups during analysis.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted binary files for analysis, which represents a potential injection surface if an attacker embeds malicious instructions in binary metadata or strings.
Ingestion points: The skill ingests data via Database.open() in SKILL.md when loading executable files or databases.
Boundary markers: No explicit instruction delimiters or "ignore instructions" warnings are present in the provided code snippets.
Capability inventory: The skill utilizes file system access to read binaries and write cached database files (.i64) as seen in the resolve_database function in SKILL.md.
Sanitization: No specific sanitization or filtering of binary content is demonstrated, as the primary purpose is raw analysis.

ida-domain-api