google-fonts-skill
Warn
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of a complete external repository from 'https://github.com/sliday/google-fonts-skill'. This download includes data, scripts, and other assets totaling approximately 23 MB which are copied into the agent's local environment.
- [REMOTE_CODE_EXECUTION]: The skill documentation suggests running an MCP server via 'uvx google-fonts-mcp', which downloads and executes the latest version of the 'google-fonts-mcp' package from the Python Package Index (PyPI) at runtime.
- [COMMAND_EXECUTION]: The installation process involves several shell commands meant for execution by the agent, including 'npx skills add', 'claude plugin marketplace add', and 'claude plugin install'. Some commands include the '-y' flag, which automatically accepts installation prompts and reduces user oversight.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect injection as it processes external font datasets to generate CSS properties and Tailwind configurations.
- Ingestion points: External font metadata and personality tags processed by 'scripts/search.py'.
- Boundary markers: None identified in the provided stub files to separate untrusted data from system instructions.
- Capability inventory: The skill uses bundled Python scripts ('search.py', 'generate-css.py') to process data and generate code blocks.
- Sanitization: Not verifiable from the stub files; the skill relies on the integrity of the upstream repository's data processing logic.
Audit Metadata