tao-finetune-cosmos-embed

Warn

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the user to execute Docker containers with broad privileges that weaken the security sandbox between the container and the host.
  • Evidence: Use of --network=host allows the container full access to host network interfaces.
  • Evidence: Use of --ipc=host allows the container to share the host's inter-process communication namespace.
  • Evidence: Elevated ulimits (memlock=-1, stack=67108864) are used to grant the container broad access to host memory resources.
  • [CREDENTIALS_UNSAFE]: The skill instructions and configuration require passing a sensitive credential (HF_TOKEN) into the container execution environment.
  • Evidence: SKILL.md and references/skill_info.yaml specify the use of -e HF_TOKEN to access gated model weights.
  • [EXTERNAL_DOWNLOADS]: The skill pulls external assets including Docker images and model weights.
  • Evidence: Resolves and pulls images from tao_toolkit.cosmos_embed.
  • Evidence: Downloads pretrained weights from nvidia/Cosmos-Embed1-* repositories on HuggingFace.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted metadata files (JSON format) and video files which could potentially contain malicious instructions targeting the processing agent.
  • Ingestion points: Untrusted JSON metadata files defined in dataset.train_dataset.metadata and video files in dataset.train_dataset.mp4_urls (SKILL.md).
  • Boundary markers: No delimiters or explicit instructions are provided to the agent to ignore embedded instructions within the processed data.
  • Capability inventory: The skill executes shell commands (Docker) and performs file system writes to the results/ directory (SKILL.md, references/skill_info.yaml).
  • Sanitization: No evidence of validation or sanitization of the input metadata content before processing.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 23, 2026, 08:15 AM
Security Audit — agent-trust-hub — tao-finetune-cosmos-embed