The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill instructs the agent to use sudo apt-get install for installing system dependencies like libreoffice and poppler-utils. Running commands with superuser privileges poses a risk to the host environment.
[COMMAND_EXECUTION]: The skill provides shell command templates for rendering spreadsheets: soffice --headless --convert-to pdf --outdir $OUTDIR $INPUT_XLSX and pdftoppm -png $OUTDIR/$BASENAME.pdf $OUTDIR/$BASENAME. If these commands are executed using unsanitized file paths or user-controlled filenames, they are vulnerable to command injection via shell metacharacters.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted data from .xlsx, .csv, and .tsv files.
Ingestion points: The skill explicitly processes tabular data from external spreadsheet files (SKILL.md).
Boundary markers: There are no instructions or delimiters provided to the agent to treat spreadsheet cell content as untrusted or to ignore embedded instructions.
Capability inventory: The skill has the capability to write to the file system and execute subprocesses via soffice and pdftoppm (SKILL.md).
Sanitization: No sanitization or validation logic is defined for the content extracted from processed spreadsheets.

spreadsheet