The Agent Skills Directory

Dynamic Execution (HIGH): scripts/office/soffice.py generates C source code at runtime, compiles it into a shared library using gcc, and injects it into the soffice process environment via LD_PRELOAD. This is a sophisticated injection technique typically scrutinized in security audits.
Command Execution (HIGH): The skill invokes several system-level binaries including gcc (for compilation), git (for diffing document text), and soffice (LibreOffice) to process documents. These operations depend on the availability and security of the host environment's toolchain.
Dynamic Code Generation (MEDIUM): scripts/accept_changes.py dynamically creates and writes a LibreOffice Basic macro to a configuration directory at runtime to automate the acceptance of tracked changes in Word documents.
Indirect Prompt Injection (LOW): The skill is designed to ingest and process external Office documents (DOCX, PPTX). While it utilizes defusedxml to protect against XML External Entity (XXE) attacks, the document content constitutes a surface for indirect prompt injection if the agent interprets text from these files as instructions.
Safe XML Handling (SAFE): The skill correctly uses the defusedxml library and lxml with appropriate caution to handle XML parsing, which mitigates standard XML-based vulnerabilities.

docx