hyperframes-registry
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill uses the
hyperframes addcommand andcurlto fetch registry metadata and component source files fromraw.githubusercontent.com/heygen-com/. This is the intended mechanism for component discovery and installation from the project's official repository. - [COMMAND_EXECUTION]: The documentation guides the agent to use various command-line tools, including the
hyperframesCLI (for adding, linting, and rendering),git(for version control), andgh(for GitHub pull requests). These tools are standard for the described development and contribution workflows. - [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and process data from external registry manifests and HTML/JS component snippets. This ingestion of untrusted external content represents a surface for indirect prompt injection.
- Ingestion points: Registry manifests (
registry.json) and component files (HTML/CSS/JS) downloaded from the remote registry URL defined inSKILL.mdand referenced inreferences/discovery.md. - Boundary markers: None identified in the instructions. The agent is directed to read and merge the content directly into the user's composition files.
- Capability inventory: File system operations (writing components), command execution (
hyperframesCLI,npxutilities), and network operations (git,gh,curl) as detailed inreferences/contributing.md. - Sanitization: No explicit sanitization or filtering of external content is mentioned, though the skill recommends running
hyperframes lintandvalidateafter integration to check for structural correctness.
Audit Metadata