The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads and transforms existing application source code, which could potentially contain malicious instructions designed to influence the agent's behavior during the analysis phase.
Ingestion points: Reading components, hooks, utility functions, application index.html, and global CSS files as described in Steps 1, 2, and 5.
Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when the agent processes the code content.
Capability inventory: The skill possesses the capability to write files to the artifacts/mockup-sandbox/ directory, execute shell commands via npm, and generate iframe previews.
Sanitization: No explicit sanitization or validation of the ingested code content is performed before it is rewritten and executed in the sandbox.
[COMMAND_EXECUTION]: The skill triggers the execution of npm run typecheck within the artifacts/mockup-sandbox/ directory to validate the integrity and correctness of the extracted components.
[EXTERNAL_DOWNLOADS]: The skill identifies and fetches configuration for fonts from Google Fonts' official domain to maintain visual fidelity in the generated mockup sandbox.

mockup-extract