gitframes
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the 'gitframes' package and related development tools from the NPM registry. It explicitly identifies that the 'skia-canvas' dependency downloads a native binary during its installation phase, which is required for the library to function correctly.
- [COMMAND_EXECUTION]: The skill provides instructions for the agent to execute shell commands to verify the Node.js environment, manage dependencies, and run video rendering tasks using scripts like 'npm run render' and 'npm run preview'.
- [DYNAMIC_EXECUTION]: The provided 'render.ts' script uses dynamic loading to import and execute composition logic from 'film.ts'. This is a standard architectural pattern for the gitframes SDK to separate video definition from rendering logic.
- [INDIRECT_PROMPT_INJECTION]: The skill establishes a workflow where external assets (specifically font files like .ttf and .otf) are ingested and processed. This presents a theoretical surface for indirect prompt injection if malicious instructions were embedded in asset metadata, although the impact is limited to the local rendering environment.
- Ingestion points: The skill reads font files from the 'assets/fonts/' directory in 'src/theme.ts'.
- Boundary markers: None specified for asset ingestion.
- Capability inventory: The skill possesses file-write capabilities (via 'fs.writeFile' in 'render.ts') and network serving capabilities (via 'startPreview').
- Sanitization: No specific sanitization or validation of font file integrity is mentioned beyond standard library loading.
Audit Metadata