The Agent Skills Directory

[SAFE]: No evidence of prompt injection or instructions to bypass safety guidelines was found. The instructions focus on structured generation of marketing content.
[SAFE]: Data retrieval is restricted to authorized workspace tools listed in the frontmatter. No attempts to access sensitive system files or exfiltrate data to unauthorized external domains were observed.
[SAFE]: There are no signs of obfuscation, hidden characters, or encoded commands within the skill's markdown files or reference libraries.
[SAFE]: The skill does not download or execute remote code, install external packages, or use dynamic execution techniques. It relies entirely on large language model reasoning and predefined tools.
[SAFE]: No persistence mechanisms, privilege escalation attempts, or time-delayed triggers were identified.

write-hooks