The Agent Skills Directory

[SAFE]: The skill provides documentation and Python code snippets for the rerun SDK. Analysis of the instructions and examples found no evidence of malicious intent, data exfiltration, or unauthorized command execution. All technical content is consistent with the stated purpose of data ingestion.
[INDIRECT_PROMPT_INJECTION]: The skill describes a pipeline for ingesting external Parquet files, which represents a theoretical attack surface for indirect prompt injection if the processed data contains malicious instructions. However, as a technical guide for developers, it does not implement an automated vulnerability, and no exploitable patterns were found in the provided examples.
[METADATA_POISONING]: The metadata fields (name, description, author) accurately reflect the content of the skill and contain no deceptive or malicious instructions.

rerun-parquet