Security audit
Security checks for vulnerabilities and agentic risk
This is a disclosed roleplay-formatting skill that changes response style for Chiikawa personas without adding code execution, data access, or persistence beyond persona-mode state.
Install this only if you want stylized Chiikawa roleplay responses. Some personas intentionally produce short or non-verbal replies, so use Disarm mode when you want the assistant to return to normal responses.
Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.
- Use short, punchy dialogue.
- Apply role-specific catchphrases, cadence, and tone.
## Name Display Rule (Language-Adaptive)
Persona names remain internal for style selection only.
Do **not** print persona names in user-facing output.
The file instructs the skill to display only one language label at runtime based on user input language rules. This imposes a locale/language behavior automatically rather than offering the user an explicit choice, which matches the policy category for forced language or locale behavior.
This is a markdown file, so SQP-2 applies to omissions in user-facing safety disclosures. The description emphasizes multi-persona roleplay and strict enforcement of communication styles, but does not warn users that outputs may intentionally become constrained, non-verbal, or character-driven in ways that could affect normal interaction expectations.
No suspicious patterns detected.