The Agent Skills Directory

[COMMAND_EXECUTION]: The skill employs common shell utilities such as grep, find, xargs, git, and npm to perform local static analysis of the project's source code and configuration files. These commands are used to identify potential security issues and are typical for auditing tasks.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data (source code, logs, and git history) to identify vulnerabilities. While the processed content could theoretically contain malicious instructions, the skill instructs the agent to report its findings sequentially rather than executing the analyzed data, which aligns with standard security auditing practices.

defense