The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted external content from research papers. 1) Ingestion points: Paper content is ingested using the read_file tool as described in references/paper-reading.md. 2) Boundary markers: The instructions do not include delimiters to isolate paper content or warnings to ignore embedded instructions. 3) Capability inventory: The skill allows for the use of write_file and edit_file to output summaries and research artifacts. 4) Sanitization: There is no validation or sanitization of the data read from external papers before processing or storage.

research-ideation