The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external data, including 'scraped' content, to generate SEO-optimized pages at scale. This ingestion of untrusted data creates a surface for indirect prompt injection attacks where malicious instructions could be embedded in the source data and potentially executed by the agent.
Ingestion points: Data requirements in SKILL.md include sourcing data from 'scraped' and 'public' sources.
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands within the processed data.
Capability inventory: The skill utilizes sandbase_call_tool to interact with SEO APIs and is capable of generating content files on the filesystem.
Sanitization: There are no explicit instructions for sanitizing or validating the ingested data before it is interpolated into templates.

programmatic-seo