The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process book annotations and highlights, which represent untrusted external data.
Ingestion points: Data is retrieved from the local Apple Books database via tools such as get_book_annotations and get_highlights_by_color.
Boundary markers: There are no instructions provided to the agent to treat highlight content as data only or to ignore potential instructions embedded within the text.
Capability inventory: The skill workflow involves writing files to the local system and executing Python scripts to process the ingested data.
Sanitization: The skill does not specify any sanitization or validation steps for the highlight text before it is used in script logic or written to the file system.
[DYNAMIC_EXECUTION]: The instructions include Python code snippets meant to be assembled and executed by the agent at runtime to overcome tool output limitations.
Evidence: The SKILL.md file contains two distinct Python blocks for programmatic color mapping and Markdown export.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of an external MCP server from a third-party source.
Evidence: The installation instructions use npx and uvx to download and configure the apple-books-mcp server from the vgnshiyer/apple-books-mcp repository.

apple-books-mcp