The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill is designed to ingest and interpret content from external data files (CSV, Excel, etc.), which represents a potential surface for indirect prompt injection if the data contains malicious instructions.
Ingestion points: Data is loaded via the DataStoryteller class from files like sales_data.csv as shown in SKILL.md.
Boundary markers: The documentation does not specify the use of delimiters or 'ignore embedded instructions' warnings for the LLM when processing row content.
Capability inventory: The skill generates narrative summaries and recommendations, implying the LLM reads and interprets data values.
Sanitization: No explicit sanitization or filtering of data content is mentioned in the provided documentation.
[Unverifiable Dependencies] (SAFE): The skill uses industry-standard Python libraries for data science (pandas, numpy, etc.). All dependencies listed in scripts/requirements.txt are well-known, versioned, and hosted on standard registries.

data-storyteller