The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill operates on external inputs such as reviewer feedback and user requests, creating a surface for indirect prompt injection. Ingestion points: Reviewer feedback and iterative refinement requests are used to guide the figure improvement process (SKILL.md). Boundary markers: No specific delimiters or ignore instructions are provided to separate user-supplied data from the agent's internal logic. Capability inventory: The skill is configured with access to the Bash tool, which allows file and directory manipulation, creating a potential path for misuse if input data is not sanitized (SKILL.md). Sanitization: There are no steps defined to sanitize or validate external content before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The documentation includes links to external resources for scientific standards and collaborative writing. References provide links to jfly.uni-koeln.de for colorblind-safe palette documentation and overleaf.com for LaTeX compilation (publication-standards.md, overleaf-packages.md).

scientific-publication