The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions were analyzed for bypass markers, developer mode overrides, or DAN-style instructions. No prompt injection attempts or safety filter bypasses were found.
[CREDENTIALS_UNSAFE]: The skill and its references were scanned for hardcoded API keys, tokens, or private credentials. No sensitive secrets were detected.
[DATA_EXFILTRATION]: No network operations (curl, wget, fetch) or sensitive file path access (e.g., .ssh, .aws) were identified. There is no evidence of data being collected or sent to external servers.
[OBFUSCATION]: The content was analyzed for Base64 encoding, zero-width characters, homoglyphs, or hidden acrostics. All text is plain and clearly readable with no hidden executable logic.
[REMOTE_CODE_EXECUTION]: The skill does not perform any external downloads or execute remote scripts. It relies entirely on static markdown instructions and local Swift code examples.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests local Swift files (**/*View.swift), which represents a potential attack surface for indirect prompt injection, it does not include dangerous capabilities such as network exfiltration or shell execution that could be exploited by embedded instructions in the code files.
[DYNAMIC_CONTEXT_INJECTION]: The SKILL.md file does not utilize any dynamic shell execution placeholders (!command).

ios-design-system