skypilot-multi-cloud-orchestration

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE] (SAFE): The skill consists entirely of documentation and configuration templates for SkyPilot. No malicious code, obfuscation, or exfiltration patterns were detected.
  • [Indirect Prompt Injection] (LOW): The skill describes how to ingest external data via YAML task definitions (task.yaml), remote Git repositories, and cloud storage (S3/GCS). While this creates an attack surface where untrusted data could influence agent actions, the documentation follows best practices for a cloud orchestration tool.
  • Ingestion points: YAML task definitions, file mounts from cloud buckets, and remote Git URLs.
  • Boundary markers: Not explicitly present in documentation snippets; relies on tool-specific parsing.
  • Capability inventory: High-privilege actions including remote shell command execution (sky launch), file synchronization, and cloud resource provisioning.
  • Sanitization: None provided in the documentation; users are responsible for validating the source of task definitions before deployment.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 04:56 PM