The Agent Skills Directory

[SAFE]: No malicious code or suspicious activities were detected. The skill consists entirely of instructional markdown and metadata defining an agentic workflow.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external content, such as draft papers and review comments, which is a necessary component of its document revision functionality. Ingestion points: Academic paper drafts and peer review feedback (referenced in the revision and revision-coach modes). Boundary markers: None explicitly defined in the provided file. Capability inventory: Generates formatted documents (LaTeX/DOCX) and visualization code (Python/R). Sanitization: No sanitization methods are described for handling the ingested text.
[DYNAMIC_EXECUTION]: The skill mentions that the Visualization Agent (Agent 11) generates Python and R code for charts. No mechanism for automatic execution of this code is provided in the skill files, classifying this as a standard feature description.

academic-paper