hyperframes-registry
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches configuration and component source code from the official HeyGen organization repository on GitHub (heygen-com/hyperframes).
- [INDIRECT_PROMPT_INJECTION]: The skill instructions involve the agent downloading and integrating external code snippets into a project, which could be exploited if a registry item were compromised.
- Ingestion points: Registry manifests and component HTML/JS files are retrieved using
hyperframes add. - Boundary markers: Not specified for the content of the registry items.
- Capability inventory: The skill has the ability to write files and execute rendering/linting commands via the hyperframes CLI.
- Sanitization: There is no mention of sanitizing or validating the content of the downloaded snippets before integration.
- [COMMAND_EXECUTION]: Employs the
hyperframesCLI tool to perform actions like adding blocks, checking project integrity, and rendering video previews. - [DATA_EXFILTRATION]: Includes a feedback mechanism that transmits search queries and missing item reports to an external endpoint to facilitate catalog improvement.
Audit Metadata