The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides the agent with access to external data from the Readwise Reader service, creating a surface for indirect prompt injection attacks.
Ingestion points: The readwise reader-get-document-details and readwise reader-list-documents commands retrieve content and metadata from documents in the user's library, which could contain instructions designed to influence the agent (SKILL.md).
Boundary markers: There are no instructions provided to the agent to use delimiters or to treat the retrieved content as untrusted data.
Capability inventory: The skill includes high-impact capabilities such as deleting documents (readwise-custom delete), creating/saving new URLs (readwise reader-create-document), and exporting the entire document library to a local ZIP file (readwise reader-export-documents).
Sanitization: No content validation or sanitization mechanisms are defined before the agent processes the retrieved document text.
[COMMAND_EXECUTION]: The skill defines an interface for executing shell commands via the readwise and readwise-custom CLI tools to perform document management tasks.
Evidence: Multiple bash commands are used for CRUD operations on documents, tags, and highlights, as well as for uploading local PDF/EPUB files (SKILL.md).

readwise-docs