bytedance-seedance-2-0
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill uses the vendor's official API endpoints (api.eachlabs.ai) and adheres to safe credential management by recommending the use of environment variables rather than hardcoding secrets.
- [EXTERNAL_DOWNLOADS]: The skill provides examples that interact with the EachLabs API and references sample resources hosted on Google Cloud Storage (storage.googleapis.com). These are recognized as legitimate vendor resources and well-known services.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing user-provided prompts and image URLs.
- Ingestion points: User-provided inputs for text prompts and starting/ending image frames as defined in SKILL.md and references/MODELS.md.
- Boundary markers: The documentation includes explicit security constraints directing the agent to validate URLs and avoid loading arbitrary content.
- Capability inventory: The skill uses curl to transmit data to the EachLabs API for processing.
- Sanitization: The skill instructions mandate validation of external URL inputs before they are utilized by the agent.
Audit Metadata