The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts scripts/generate_pdf.py and scripts/generate_schematic.py utilize subprocess.run to execute system binaries like pandoc and xelatex, as well as internal Python scripts. These operations are necessary for document formatting and image generation. Commands are executed using argument lists (shell=False), which follows security best practices to prevent argument injection.
[EXTERNAL_DOWNLOADS]: The skill performs network requests to well-known and reputable services for its core functionality. scripts/generate_schematic_ai.py communicates with the OpenRouter API for image generation, while scripts/verify_citations.py interacts with doi.org and api.crossref.org to retrieve citation metadata. These operations are consistent with the skill's purpose of academic research and synthesis.
[PROMPT_INJECTION]: The skill is designed to process external research data, creating a surface for indirect prompt injection.
Ingestion points: Untrusted content from academic databases (e.g., titles and abstracts) is ingested via scripts/search_databases.py and scripts/verify_citations.py.
Boundary markers: Absent. External data is not wrapped in specific delimiters within the assets/review_template.md to distinguish it from system instructions.
Capability inventory: The skill can execute shell commands for PDF generation (scripts/generate_pdf.py) and coordinate sub-processes for image generation (scripts/generate_schematic.py).
Sanitization: Absent. The skill aggregates data into markdown format without explicit sanitization or escaping of potential injection patterns found in external records.

literature-review