cdf-image-attribute-augmentation
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
kubectl,curl, andawsCLI tools to interact with the user's Kubernetes cluster, Airflow API, and S3 storage. These operations are core to its functionality for orchestrating DAG runs. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data (person-crop images from S3 and local paths) and interacts with LLM/VLM endpoints. While it documents path validation in
scripts/workflow.pyandscripts/upload_images.py, the ingestion of untrusted data into an agent-managed workflow represents an attack surface. The risk is mitigated by explicit instructions to not invent endpoints and to use strict payload validation.
Audit Metadata