The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted external content, such as website URLs and source code, to perform design audits. This ingestion path creates a surface for indirect prompt injection where malicious instructions could be embedded within the design data being analyzed.
[COMMAND_EXECUTION]: The skill's configuration includes the Bash tool in its allowed-tools list. While no malicious shell commands are present in the provided instructions, the availability of a shell environment alongside the processing of external data represents a capability that requires the agent to maintain strict adherence to its primary task.

form-exam