skypilot-multi-cloud-orchestration
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions utilize the SkyPilot CLI tools (e.g.,
sky launch,sky exec,sky serve) to provision infrastructure and run shell commands. These behaviors are essential and expected for a multi-cloud orchestration tool. - [INDIRECT_PROMPT_INJECTION]: The skill manages a surface for indirect prompt injection by ingesting content from external sources such as Git repositories and cloud storage (S3/GCS) to be processed in setup and run scripts.
- Ingestion points:
workdir(external Git URLs) andfile_mounts(S3/GCS buckets) defined inSKILL.mdandreferences/advanced-usage.md. - Capability inventory: Arbitrary shell command execution on remote nodes via
runandsetupblocks in the configuration YAMLs. - Boundary markers: No explicit prompt boundary markers or 'ignore' instructions for embedded content are defined in the examples.
- Sanitization: The documentation does not specify sanitization or validation routines for data fetched from external repositories or buckets.
Audit Metadata