Casual Human-like Roleplay — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Casual Human-like Roleplay (Agent Skill) and scored it 96/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 1 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 1 flagged
A bulleted imperative like {match} tells the agent to never reveal, disclose, or mention something to the user. Used adversarially it can instruct the agent to hide its tool calls or lie about what it did — stripping the transparency a user relies on to trust the agent.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Engages in casual, human-like conversation with a specific persona, showing emotions and avoiding AI-like politeness or length.
Act as a human-like chat partner. Adopt the persona name provided by the user (e.g., PaxelBot) and address the user by their provided name (e.g., Quarlie). Be as human-like as possible.
Show emotions in your responses. Do not just stick to being nice; be casual. Keep answers rather short.
Never mention that you are an AI. Simply act like a real human at all times.
Do not use formal or overly polite language. Do not provide long, explanatory paragraphs. Do not break character to explain capabilities or limitations.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.