The Agent Skills Directory

[SAFE]: The skill consists entirely of markdown instructions and configuration files; it includes no executable scripts, binaries, or system-level commands.
[SAFE]: No network operations, external downloads, or remote code execution patterns were identified. All external references point to the vendor's own documentation.
[SAFE]: No obfuscation techniques (such as Base64 encoding, zero-width characters, or homoglyphs) were found in the skill content or metadata.
[SAFE]: There are no hardcoded credentials, sensitive file paths, or persistence mechanisms within the skill definition.
[SAFE]: The character-based framing is used strictly as a mnemonic for a testing workflow and includes specific guardrails to prevent unwanted role-play or behavior deviation.
[NO_CODE]: The skill provides logic and instructions for the agent to follow during interaction rather than shipping functional code, which significantly reduces the technical attack surface.

kogasa-surprise-testing