The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions facilitate the ingestion and analysis of external, untrusted scholarly documents, creating an indirect prompt injection surface.
Ingestion points: Research papers, proposals, and abstracts are ingested into the agent context for evaluation as described in the workflow in SKILL.md.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands within the evaluated documents are specified in the workflow.
Capability inventory: The skill performs local file-write operations to save reports via scripts/calculate_scores.py and references a schematic generation capability via scripts/generate_schematic.py.
Sanitization: There is no evidence of sanitization or filtering applied to the content of the analyzed documents.
[COMMAND_EXECUTION]: The skill provides instructions for the agent to execute local Python scripts provided in the skill folder. The script scripts/calculate_scores.py is used to automate the scoring of research dimensions and save reports to the local file system. The script logic is transparent and uses standard library functions.
[EXTERNAL_DOWNLOADS]: The skill references the ScholarEval framework methodology hosted on arxiv.org. This is a well-known and reputable scientific repository for research preprints used for identifying evaluation criteria.

scholar-evaluation