The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's metadata (allowed-tools) explicitly requests access to the shell tool, which allows for the execution of arbitrary system commands.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from external files (company-profile.md and HUMAN_AGENDA.md) as part of its operational context. This configuration presents a vulnerability surface where instructions embedded in those external files could influence the agent's behavior, potentially leveraging the requested shell capabilities.

ceo