The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill requires the agent to fetch and analyze content from public-facing web pages, which is an untrusted data source. This creates a vulnerability where malicious instructions hidden in a website could attempt to influence the agent's actions during the audit. Ingestion points: As defined in Phase 1 and Phase 2 of 'SKILL.md', the agent is directed to retrieve external files (e.g., '/robots.txt', '/sitemap.xml') and page content (e.g., titles, meta tags, and structured data). Boundary markers: The instructions do not provide explicit delimiters or instructions for the agent to ignore potentially malicious content within the audited data. Capability inventory: The workflow involves network operations to fetch target pages and file modification capabilities to apply SEO fixes (Phase 3), providing a mechanism for an injection to trigger unauthorized actions. Sanitization: The skill does not specify any validation or sanitization routines for the external data retrieved from the target websites.

seo-audit