The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data through user-provided answers to a discovery questionnaire (e.g., product descriptions, target audience details, and competitor information). This data is interpolated into templates to generate advertorial copy. However, the skill lacks any capabilities for command execution, file system access, or network operations, meaning any potential injection would only affect the generated text output.
[SAFE]: The skill does not include any executable scripts, binaries, or platform-level configuration that would enable tool invocation or system access.
[SAFE]: No signs of obfuscation, Base64 encoding, homoglyphs, or hidden/encoded URLs were found within the instructions or metadata.
[SAFE]: All referenced platforms (Taboola, Outbrain, Facebook) and expert names (Russell Brunson, Neil Patel, etc.) are used in their standard professional context for marketing advice and do not represent malicious redirects or impersonation attempts.

advertorial-writing