roboflow-batch-processing

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill allows the agent to process external data sources including local directories, cloud storage buckets, and reference files, which represents a potential surface for indirect prompt injection.
  • Ingestion points: Data is ingested from local paths via --images-dir and --videos-dir, cloud storage via --bucket-path, and JSONL files via --references in SKILL.md.
  • Boundary markers: The skill does not specify any boundary markers or instructions to ignore embedded commands within the processed data.
  • Capability inventory: The skill involves shell command execution through the inference-cli and network communication for staging and job status in SKILL.md.
  • Sanitization: No explicit sanitization or filtering of external content is mentioned.
  • [EXTERNAL_DOWNLOADS]: The instructions recommend installing the inference-cli package, which is a standard library provided by the vendor for their batch processing service.
  • [COMMAND_EXECUTION]: The skill provides numerous shell commands for the inference-cli to perform data staging, job management, and log retrieval.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 02:54 AM
Security Audit — agent-trust-hub — roboflow-batch-processing