The Agent Skills Directory

[NO_CODE]: The skill consists entirely of Markdown instructions and logic frameworks. There are no executable scripts, binaries, or configuration files included in the skill package.
[DATA_EXFILTRATION]: The skill explicitly forbids automated responses ('Never auto-send') and does not include any network or file-system capabilities that would allow for unauthorized data transfer.
[SAFE]: The instructions implement several security best practices, including a 'Quality Bar' that requires all claims to be sourced from specific profile proof points or journalist context, and a 'Decline/Ethics gate' that prevents the agent from fabricating expertise or responding to out-of-lane queries.

reactive-comment