data-pipelines
Pass
Audited by Gen Agent Trust Hub on Feb 24, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: Indirect prompt injection surface identified through the skill's core data ingestion functionality.
- Ingestion points: External data is brought into the context via dlt (REST API source) and DuckDB's file reading capabilities (CSV, Parquet, and S3).
- Boundary markers: The skill does not instruct the agent to use delimiters or specific ignore instructions when processing data from these external sources.
- Capability inventory: The skill provides mechanisms for executing Python via uv run and SQL transformations via sqlmesh and DuckDB, which can be influenced by ingested data.
- Sanitization: No sanitization or schema validation rules are explicitly defined for the data extracted from remote APIs or external files.
- [COMMAND_EXECUTION]: Documentation of standard command-line tools for workflow automation.
- Evidence: The skill details the use of uv run for executing Python scripts and managing dependencies, along with CLI usage for dlt, sqlmesh, and DuckDB, which are typical for data engineering tasks.
Audit Metadata