The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external PDF files, creating a surface for indirect prompt injection.
Ingestion points: The skill uses pypdf, pdfplumber, pdftotext, and OCR via pytesseract to read text content from PDF files into the agent's context.
Boundary markers: Absent. No instructions are provided to the agent to treat extracted PDF content as untrusted or to use specific delimiters when processing the text.
Capability inventory: The skill has the capability to write files to the local system and execute command-line utilities like qpdf, pdftk, and pdftotext.
Sanitization: Absent. Extracted text is used directly without filtering or validation against potential malicious instructions embedded in the documents.
[EXTERNAL_DOWNLOADS]: The skill references the installation of standard, well-known third-party libraries.
Evidence: Mentions pip install pytesseract pdf2image in SKILL.md. These are established packages for OCR and PDF image conversion.
[COMMAND_EXECUTION]: The skill provides several command-line examples for common PDF utilities.
Evidence: Includes usage patterns for pdftotext, qpdf, pdftk, and pdfimages to perform operations like text extraction, merging, and decryption.

pdf