The Agent Skills Directory

[DATA_PROCESSING]: The skill performs legitimate data transformation operations using the pandas library. It supports multiple input formats (CSV, Excel, JSON) and converts them to Parquet for optimized storage and querying.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes untrusted external data files.
Ingestion points: Data is ingested through pd.read_csv, pd.read_excel, pd.read_json, and pd.read_parquet in SKILL.md.
Boundary markers: None identified in the prompt templates.
Capability inventory: The skill is limited to filesystem read/write operations and data structure manipulation. It does not possess high-risk capabilities like network access or arbitrary command execution.
Sanitization: There is no explicit sanitization of the content within the data files before the agent presents results to the user.
[FILESYSTEM_ACCESS]: The skill correctly declares filesystem permissions in claw.json, which is necessary for its stated purpose of converting local data files.

parquet-converter