The Agent Skills Directory

[COMMAND_EXECUTION]: The skill requires the execution of shell commands to generate scientific diagrams via Python scripts and to compile LaTeX documents using xelatex and bibtex.
[EXTERNAL_DOWNLOADS]: The skill instructions guide the agent to perform web searches and fetch literature from external sources such as PubMed. Additionally, the included scripts (scripts/generate_schematic_ai.py) make network requests to the openrouter.ai API for image generation and review.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface (Category 8) as it processes untrusted scientific literature from the web and integrates it into the agent's reasoning and final report generation. There are no explicit boundary markers or sanitization steps implemented for this external data.
Ingestion points: Scientific literature fetched via WebFetch (PubMed) and WebSearch as described in the workflow sections of SKILL.md.
Boundary markers: Absent; the skill does not use specific delimiters or warnings to separate fetched literature from core instructions.
Capability inventory: The skill uses Write, Edit, and Bash tools, and executes subprocesses in scripts/generate_schematic.py and scripts/generate_schematic_ai.py.
Sanitization: No validation or filtering is performed on the content retrieved from external literature sources.

hypothesis-generation