The Agent Skills Directory

[PROMPT_INJECTION]: The skill describes a workflow that ingests data from external sources, creating an inherent risk for indirect prompt injection.
Ingestion points: The skill uses web_search to find learning resources and read_file to analyze study materials.
Boundary markers: There are no instructions or delimiters specified to differentiate external data from agent instructions or to ignore commands embedded in the retrieved content.
Capability inventory: The skill integrates several tools including web_search, note, calendar_event, and read_file.
Sanitization: The skill does not mention any sanitization, filtering, or validation of the content processed from external tools.
[NO_CODE]: The skill consists exclusively of instructional markdown content and metadata. It does not package any scripts or binary files, which prevents direct code execution or traditional malware behavior.

study-helper