The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx -y @sjunepark/fs@0.2.0 to download and execute a package from the public npm registry. This occurs during schema loading, document validation, and rendering steps.
[COMMAND_EXECUTION]: Several shell commands are executed using npx to interface with the financial statement tool, including schema document, example, validate, render, and create. These commands process local files and generate new outputs.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it ingests and resolves untrusted data from varied external formats like workbooks, PDFs, images, and webpages. * Ingestion points: Source materials such as spreadsheets, PDFs, and OCR transcripts are read into the agent's context to resolve the financial model. * Boundary markers: There are no explicit delimiters or instructions to the agent to ignore potentially malicious embedded content within the source documents. * Capability inventory: The agent has the ability to write files (JSON and HTML) and execute shell commands via subprocesses. * Sanitization: The instructions do not define any sanitization or filtering procedures for data extracted from external artifacts.

author-fs