The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides patterns for processing data from external streaming sources (Kafka), creating a potential attack surface if the stream content is malicious.
Ingestion points: Data ingestion from Kafka topics is shown in references/kafka-streaming.md.
Boundary markers: No specific delimiters are included in the provided code templates for the processed data.
Capability inventory: The skill demonstrates file and table writes using writeStream.start and saveAsTable in SKILL.md and references/multi-sink-writes.md.
Sanitization: Patterns use from_json with a defined schema to validate and structure the incoming data.
[DYNAMIC_EXECUTION]: Code snippets use string interpolation to construct Spark SQL statements and file paths.
Evidence: Dynamic construction of MERGE queries in references/merge-operations.md and table paths in references/checkpoint-best-practices.md.
[COMMAND_EXECUTION]: The skill includes instructions for managing the Databricks filesystem using dbutils.fs commands.
Evidence: Use of dbutils.fs.rm, dbutils.fs.cp, and dbutils.fs.ls for checkpoint management and recovery in references/checkpoint-best-practices.md.

databricks-spark-structured-streaming