The Agent Skills Directory

[SAFE]: The skill provides instructional content and SQL templates for performing static analysis on binary files. The behavior is consistent with the stated purpose of analyzing radare2 cross-references via the r2xsql interface.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, potentially untrusted sources (compiled binaries being analyzed).
Ingestion points: Data enters the context from radare2's analysis state through r2xsql table queries (e.g., xrefs, strings).
Boundary markers: No explicit boundary markers or instructions to ignore embedded content are defined in the provided SQL templates.
Capability inventory: The skill uses the Bash and Read tools to interact with the underlying analysis engine.
Sanitization: The skill utilizes SQL for structured data retrieval, which inherently provides a layer of data isolation compared to raw text processing, though the resulting strings from the binary (like string_value) are passed to the agent.

xrefs