The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of prompt injection or attempt to override agent safety guidelines was found. The use of markers like [CRITICAL] and [HIGH] in the text refers to the priority of web accessibility standards, not instructions for the AI agent to bypass constraints.
[DATA_EXPOSURE_AND_EXFILTRATION]: No sensitive file paths, hardcoded credentials, or unauthorized network communication patterns were detected. The JavaScript snippets provided are generic educational examples (e.g., modal focus traps, route splitting) and do not perform any data exfiltration.
[OBFUSCATION]: The content is written in clear, human-readable Markdown and code. No Base64, zero-width characters, homoglyphs, or other obfuscation techniques were identified.
[REMOTE_CODE_EXECUTION]: The skill does not instruct the agent to download or execute external scripts or packages. Reference to resource hints (e.g., preconnecting to Google Fonts) follows standard web development best practices.
[PRIVILEGE_ESCALATION]: No commands related to privilege escalation (e.g., sudo, chmod) are present.
[PERSISTENCE_MECHANISMS]: The skill does not attempt to establish persistence on the host system.
[METADATA_POISONING]: The metadata fields are used appropriately to describe the skill's purpose and author, with no deceptive or malicious content.
[INDIRECT_PROMPT_INJECTION]: The skill does not ingest untrusted external data and lacks attack surfaces for indirect injection.
[DYNAMIC_EXECUTION]: No evidence of dynamic code generation, unsafe deserialization, or library injection was found.
[DYNAMIC_CONTEXT_INJECTION]: The skill does not utilize dynamic shell command execution placeholders.

web-design-guidelines