The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill utilizes tools that generate and execute JavaScript at runtime to perform complex layout and design tasks.
Evidence:
references/edit-code.md defines moda canvas edit as a 'sandboxed JS batch editor' that runs synchronous JavaScript against a live canvas.
references/markup.md describes a <generate> tag used within XML markup that executes JavaScript to emit elements.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from untrusted external sources, creating a potential surface for indirect prompt injection.
Ingestion points: The workflow described in references/multi-unit-workflow.md suggests using moda web read <url> to extract palettes and styles from third-party websites.
Boundary markers: There are no specific instructions or delimiters mentioned to prevent the agent from following instructions embedded in the content fetched from external URLs.
Capability inventory: The skill possesses significant capabilities, including the ability to write/modify canvas content, execute JavaScript in a sandbox, and generate media via moda media.
Sanitization: While the JavaScript execution environment is described as a sandbox with limited APIs and resource constraints, there is no specific sanitization mentioned for the natural language content processed from web research.

moda-social-youtube