The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to manage files and directories.
Evidence: Employs find to locate documents, mkdir -p to create folder structures, and cp or mv to organize files based on extracted metadata (SKILL.md).
[DATA_EXFILTRATION]: The skill accesses and aggregates sensitive financial data, which constitutes a data exposure risk.
Evidence: The skill is designed to read and extract vendor names, invoice numbers, dates, amounts, and payment methods from private financial documents (SKILL.md).
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it ingests untrusted data from external files.
Ingestion points: The agent is instructed to read text and metadata from PDF invoices and image-based receipts provided by the user (SKILL.md).
Boundary markers: Absent. The instructions do not include delimiters or warnings for the agent to ignore potentially malicious instructions embedded within the invoice text.
Capability inventory: The skill can perform directory traversal (find), create directories (mkdir), move/copy files (mv/cp), and generate summary reports (CSV file creation) (SKILL.md).
Sanitization: Absent. There are no instructions to sanitize or validate the text extracted from documents before using it for file operations or report generation.

invoice-organizer