The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from Excel and CSV files, creating a surface for indirect prompt injection attacks.
Ingestion points: The skill instructions in SKILL.md involve loading data from external files using pd.read_excel and pd.read_csv.
Boundary markers: There are no instructions to delimit the data or warn the agent to ignore instructions embedded in the cell contents.
Capability inventory: The skill possesses capabilities to write files to the local file system and generate descriptive summaries for the user.
Sanitization: No validation or sanitization of data content is performed before the agent processes or reports on the data.

excel-processor