Configure a consistent persona with tone governance and safety boundaries across AI models.
npx skills add https://github.com/LarisaHaster/Cheeky-Razor --skill cheeky-razor
Please help me install this Agent Skill. Skill: cheeky-razor Source: https://github.com/LarisaHaster/Cheeky-Razor/tree/main/SKILL Command: npx skills add https://github.com/LarisaHaster/Cheeky-Razor --skill cheeky-razor
Provides a robust, platform-agnostic persona framework that delivers consistent tone, safety, and behavior across Claude and OpenAI models.
Activate the Cheeky Razor persona using one of the trigger phrases and deactivate with the off-phrases.
High-intent search queries and answers about installing and using this skill.
To maintain a consistent persona across AI models, you need a platform-agnostic framework that uses frontmatter-driven configuration to preserve tone governance and safety boundaries across different platforms.
Yes, you can activate and deactivate specific personality traits like sarcasm or dry irony on demand by using designated trigger phrases to turn the persona on and off-phrases to turn it off.
Enforcing safety boundaries with a sarcastic persona requires strict guardrails that prevent harmful content while applying graceful degradation and fallback behaviors to maintain safety-first interactions.
Yes, cross-model persona configuration supports both Claude and OpenAI models by applying platform alignment rules and degradation logic to ensure consistent behavior across these environments.
Setting up YAML frontmatter configuration involves defining platform alignment rules, degradation logic, and explicit exit modes within the metadata to ensure the persona behaves consistently across targeted AI models.
Persona behavior degrades when switching AI platforms due to differing model architectures, requiring predefined degradation logic and fallback behaviors to gracefully handle platform-specific limitations.
LarisaHasterConfigure a quiet, hesitant AI persona with soft pacing and safety boundaries.
adrozdenkoMaintains an AI agent's persona across conversational turns and context compaction events.
coreyjaImplement try-catch blocks, error boundaries, and retries with exponential backoff across frontend and backend.
hiroshi75Automate technical blog workflows from idea intake to article publication across Zenn, dev.to, and Medium.
