The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and parses untrusted data from external websites during the 'Fetch & Parse' step. Maliciously crafted content on the target page could attempt to influence the agent's behavior or bias its report.
Ingestion points: Step 1: Fetch & Parse in SKILL.md captures full rendered page content, structure, and CTAs.
Boundary markers: The skill lacks instructions to the agent to treat the fetched content as data only or to ignore embedded instructions (e.g., delimiters like XML tags or triple backticks with warnings).
Capability inventory: The skill is primarily analytical; it generates markdown reports and does not possess capabilities for file system modification, command execution, or network exfiltration of sensitive local data.
Sanitization: There are no specified sanitization or validation steps for the fetched content before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill requires internet access to fetch external web content for analysis. This is a core functionality necessary for the skill's stated purpose of auditing landing pages and is explicitly declared in the frontmatter.

conversion-audit