skills/datadrivenconstruction/ddc_skills_for_ai_agents_in_construction/parquet-converter/Gen Agent Trust Hub
parquet-converter
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data files (CSV, Excel, JSON, Parquet) which acts as an attack surface for indirect prompt injection.
- Ingestion points: Data is ingested via
pd.read_csv,pd.read_excel,pd.read_json, andpd.read_parquetinSKILL.md. - Boundary markers: The skill does not implement explicit boundary markers or specific 'ignore instructions' prompts to isolate data content from the agent's control logic.
- Capability inventory: The skill allows file system read and write operations via the
pandasandpyarrowlibraries to facilitate data conversion and logging. - Sanitization: No sanitization is performed on the data content itself to filter out potential natural language instructions before the agent interprets or summarizes the results.
Audit Metadata