The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, untrusted sources such as URLs, PDFs, and articles.
Ingestion points: The workflow in SKILL.md fetches content from user-provided URLs or files using the fetch-content tool or standard web fetchers.
Boundary markers: The instructions lack explicit delimitation or specific "ignore embedded instructions" warnings for the fetched content.
Capability inventory: No high-risk capabilities, such as file-system writes or arbitrary command execution, are present in this skill.
Sanitization: No sanitization or validation of the fetched content is specified before the summarization process begins.

summarize