The Agent Skills Directory

[SAFE]: The skill contains legitimate technical documentation, best practices, and code templates for building Astro components. No malicious patterns such as obfuscation, exfiltration, or unauthorized command execution were found.
[INDIRECT_PROMPT_INJECTION]: The skill addresses potential injection surfaces by providing best practices for data passing.
Ingestion points: Components ingest data via Astro props (references/prop-serialization.md) and user input via the search template (references/templates/interactive-island.md).
Boundary markers: The documentation explicitly defines serialization rules to maintain boundaries between server and client contexts.
Capability inventory: Interactive templates include client-side network requests (fetch).
Sanitization: The code examples demonstrate safe handling of untrusted data using encodeURIComponent to sanitize query parameters.

astro-islands