The Agent Skills Directory

[SAFE]: The skill functions as a documentation and skill generator for data analysis workflows, intended to help users build localized context for their specific data warehouses.
[COMMAND_EXECUTION]: The skill generates SQL discovery queries (e.g., querying INFORMATION_SCHEMA.SCHEMATA) to identify database structures. These queries are standard for the tool's stated purpose and do not represent malicious command execution.
[INDIRECT_PROMPT_INJECTION]: The skill contains an attack surface for indirect prompt injection as it processes external data (database schemas and user responses) to generate new instructions.
Ingestion points: Data enters the context via database schema tool outputs and user answers to discovery questions.
Boundary markers: None identified; the skill does not explicitly provide delimiters to separate potentially malicious schema names or user-provided definitions from the generated skill's instruction set.
Capability inventory: The skill uses Read, Glob, and Grep, and it has the capability to write local files and package them into a ZIP archive.
Sanitization: No validation or sanitization of ingested schema metadata or user responses is performed before they are incorporated into the generated templates.
Note: This is an inherent risk factor of a documentation-generation tool and is considered acceptable within its primary use-case.

data-context-extractor