The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by reading local repository files to determine the application entrypoint and package versions, which creates a surface where malicious content in those files could influence the agent's behavior.
Ingestion points: The skill uses Read, Glob, and Grep tools to inspect files within the user-provided [project-path].
Boundary markers: There are no explicit instructions or delimiters defined to isolate the data read from the files from the agent's operating instructions.
Capability inventory: The skill has the ability to Write and Edit files, and perform network requests using WebFetch.
Sanitization: No validation or sanitization of the content retrieved from the local files is specified before it is used to plan the smoke test.
[DYNAMIC_EXECUTION]: The skill involves the dynamic generation and execution of code within the WebContainer browser runtime.
This includes instructions for WebContainer.boot(), mounting a FileSystemTree, and using spawn() to run installation and server processes.
[EXTERNAL_DOWNLOADS]: The skill references external documentation and API resources from webcontainers.io.
These references target a well-known service related to the skill's core purpose and are used for guidance rather than direct code execution.

stackblitz-hello-world