The Agent Skills Directory

[SAFE]: The skill's primary purpose is SEO auditing and site planning, and all instructions and references align with this goal.
[PROMPT_INJECTION]: No instructions were found that attempt to override safety filters or extract system prompts. The skill processes external site content (via web_fetch or curl) and user-provided HTML, which is a standard surface for its SEO auditing function. Ingestion points include external site content, user-provided HTML/screenshots, and brand configuration files. Boundary markers are absent, and there is no explicit sanitization logic for external content, but the capability is limited to file write access in the marketing/seo/ directory.
[DATA_EXFILTRATION]: The skill interacts with user-specified websites for auditing purposes. References to validation services target well-known and established domains (Google, Schema.org). No sensitive local file access or unauthorized data transfer patterns were identified.
[REMOTE_CODE_EXECUTION]: The skill does not download or execute remote scripts or binaries.
[COMMAND_EXECUTION]: The skill uses curl to fetch website data, which is an expected tool for its technical audit function. No signs of command injection or malicious shell command usage were found.
[OBFUSCATION]: The skill uses plain-text instructions and references, with no detection of hidden characters or encoded strings.

seo-audit