hiapi-seedance-2-0-video
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches a version manifest from
https://raw.githubusercontent.com/HiAPIAI/hiapi-skills/main/skills.jsonto check for available updates. - [COMMAND_EXECUTION]: The installer script
scripts/install.mjsutilizeschild_process.execSyncto executegit clonefor downloading the skill's source code. - [REMOTE_CODE_EXECUTION]: Installation instructions suggest using
npxto run code directly from the vendor's GitHub repository (npx -y github:HiAPIAI/hiapi-seedance-2-0-video-skill), which involves the execution of remote scripts. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-controlled input, such as text prompts and media URLs, which are subsequently sent to the remote HiAPI service.
- Ingestion points: User-supplied arguments
--prompt,--first-frame-url, and other media flags inscripts/hiapi-seedance-2-video.mjs. - Boundary markers: None identified in the prompt construction.
- Capability inventory: Performs network requests via
fetchand writes files to the localoutputs/directory usingfs/promises.writeFile. - Sanitization: Implements basic parameter normalization for technical constraints (e.g., resolution, duration) in
scripts/lib/seedance-2-video.mjs, though it does not sanitize the natural language prompt content.
Audit Metadata