The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the paperbanana[mcp] package via pip and its execution via uvx. These resources are fetched from the public PyPI registry.
[COMMAND_EXECUTION]: The skill's functionality is delivered through a command-line tool (paperbanana). The agent is instructed to execute various subcommands such as generate, plot, evaluate, and batch to perform figure creation and assessment.
[INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting untrusted external data, including research paper PDFs, methodology text files, and CSV datasets. This data is interpolated into prompts for a multi-agent pipeline, creating a surface for indirect prompt injection attacks where malicious instructions hidden in the documents could attempt to influence agent behavior.
[DYNAMIC_EXECUTION]: The plot command generates visualization code (specifically Matplotlib) based on input data and intent descriptions. This code is then executed to produce the requested statistical charts.

paper-banana