The Agent Skills Directory

[NO_CODE]: The skill consists entirely of Markdown-based guidance and workflow instructions. No Python scripts, Node.js modules, Shell scripts, or binary files are present in the skill directory.
[PROMPT_INJECTION]: The skill defines an indirect prompt injection surface through its research instructions in modules/research-checklist.md.
Ingestion points: The instructions direct the agent to perform web research and use tools to distill information from Wikipedia and Arxiv into its knowledge base.
Boundary markers: Absent. There are no instructions to use delimiters or warnings to ignore embedded instructions within the retrieved external data.
Capability inventory: The agent provides expert scientific advice; successful poisoning of the research sources could lead to the dissemination of misinformation or biased content.
Sanitization: Absent. The skill does not provide any criteria for validating or filtering the retrieved data before it is integrated into the agent's responses.

neuroscience-fundamentals