The Agent Skills Directory

[SAFE]: The skill instructions and reference materials provide best-practice guidance for building server-side templates. It explicitly emphasizes security measures such as using escaped output {{ }} by default to mitigate XSS risks and provides clear warnings for using unescaped output {{{ }}} only with trusted, sanitized content.
[SAFE]: Documentation for form components includes built-in CSRF protection mechanisms and method spoofing helpers, promoting secure web development patterns within the AdonisJS ecosystem.
[INDIRECT_PROMPT_INJECTION]: While the skill defines a surface for processing untrusted data (template rendering), it provides robust instruction on boundary markers and sanitization.
Ingestion points: Data passed to view.render() and HTTP request inputs via the request helper.
Boundary markers: Uses double-curly {{ }} delimiters for automatic escaping.
Capability inventory: No dangerous shell execution or network exfiltration capabilities are present in the skill's own code or recommended patterns.
Sanitization: Built-in HTML escaping is the default behavior described and enforced throughout the documentation.

edge