The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/office/soffice.py dynamically generates C source code, compiles it into a shared library using gcc at runtime, and uses LD_PRELOAD to inject the library into the LibreOffice process. This mechanism is intended to allow LibreOffice to function in restricted or sandboxed environments.
[COMMAND_EXECUTION]: The skill invokes external command-line utilities via the subprocess module:
gcc is called to compile technical shims in scripts/office/soffice.py.
soffice is executed to recalculate spreadsheet formulas in scripts/recalc.py.
git is used for comparing document differences in scripts/office/validators/redlining.py.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: Processes user-supplied spreadsheet files (.xlsx, .csv, .tsv) through various scripts.
Boundary markers: Instructions do not define markers for separating user data from agent instructions.
Capability inventory: The skill can write to the file system, execute shell commands, and run LibreOffice macros for recalculation.
Sanitization: Uses defusedxml to protect against XML-based vulnerabilities during document processing.

xlsx