spark
Pass
Audited by Gen Agent Trust Hub on Mar 7, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection through its data ingestion capabilities.
- Ingestion points: The skill provides patterns for reading data from external sources including HDFS (spark.read.text), Amazon S3 (s3://bucket/data.csv), and Kafka streams (spark.readStream).
- Boundary markers: The provided PySpark examples do not utilize delimiters or instructions to ignore instructions embedded within the processed data.
- Capability inventory: The skill includes scripts with capabilities to write data to the filesystem (saveAsTextFile), Amazon S3, and external databases via JDBC (write.jdbc).
- Sanitization: No input validation, escaping, or sanitization logic is present in the examples for handling content retrieved from external sources.
Audit Metadata