The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from external sources such as keyword exports, customer call transcripts, survey responses, and forum posts (Reddit, Quora).
Ingestion points: The instructions in SKILL.md define processes for analyzing external SEO data, call transcripts, and web search results.
Boundary markers: The skill does not provide explicit delimiters or "ignore embedded instructions" warnings for the processed data.
Capability inventory: The skill uses text generation and web search tools; it does not utilize file-write, network exfiltration, or arbitrary code execution capabilities.
Sanitization: There are no mentioned sanitization or validation steps for the external data being processed.
[SAFE]: The skill references a local configuration file .claude/product-marketing-context.md to gather business context, which is a standard practice for maintaining project-specific information within the agent environment and does not involve sensitive system files or credentials.

content-strategy