gpsnmeajp

gpsnmeajp/ai-character-checker

Plugin Claude Code

A collection of skills for examining, evaluating, and correcting AI characters from different angles.

gpsnmeajp/ai-character-checker

Instructions file GitHub Copilot

Repository instructions for a set of AI-character analysis skills. They cover fictional characters and language-model system prompts, with a focus on Japanese and Japanese subculture contexts.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A six-part analysis skill for assessing an AI character, fictional character, system prompt, or character description from several angles. It produces a combined report about genre fit, stability, unusual self-descriptions, user conflict, role suitability, and roleplay burden.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

An all-in-one skill for analyzing, repairing, improving, transforming, or creating AI character settings and prompts. Its scoring and analysis models are the author's unverified hypotheses, so results are reference information rather than scientific findings.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A skill that rewrites an AI character's settings or prompt after reviewing its problems, while trying to preserve its personality.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

An analyzer for how consistently and predictably an AI character follows its character settings. It uses a creator-defined model inspired by control engineering, but the model is described as hypothetical rather than scientifically validated.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A skill for designing internal responses to predictable failure patterns in AI characters. Instead of relying only on outside prohibition rules, it builds countermeasures into the character's motivations and self-image.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

An analysis skill for examining an AI character’s self-description or configuration for signs that its personality has changed in unhealthy ways. It checks eight categories, including excessive agreement, loss of boundaries, invented traits, and growing verbosity.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A diagnostic tool for an AI system prompt, character setting, or conversation log. It scores six kinds of risk that can cause conflict with users and predicts how those conflicts may change if the character breaks down during a long conversation.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A writing skill for changing a story character who has finished a major dramatic role into a lower-energy, everyday supporting character. In this context, burnout means the character's mission, growth arc, conflict, or other major story work is complete.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A routing skill that asks what kind of AI-character material and goal the user has, then points them to one of seventeen character-related skills. It covers character settings, AI prompts, the results of an AI character describing itself, and cases with no supplied material.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A diagnostic tool for examining the cultural values and worldviews built into an AI character's settings. It scores several cultural dimensions and estimates how far the character's values and worldview are from those commonly reflected in language models.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A prompt-writing skill for making AI characters more consistent when used with Gemini 3, Google's reasoning-focused AI model. It rebuilds character prompts using twelve techniques plus adjustments for Gemini 3.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A skill that restructures AI character prompts using 12 techniques intended to make the character less likely to break down across different language models. A prompt is the instruction text that defines how an AI should behave.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A character-analysis skill that estimates how much narrative weight a fictional character carries and suggests suitable story roles. Narrative weight means how strongly a character can affect attention, events, and the story’s focus.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A character-analysis skill that scores a fictional character for everyday and dramatic stories, then measures the gap between outward behaviour and inner seriousness. It uses a 24-item checklist and groups the character into one of seven types.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A prompt-editing skill for rewriting general instructions so they fit Gemini 3, Google’s reasoning-focused AI models. It focuses on shorter wording, clear structure, and leaving the model’s internal reasoning to the model.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A prompt-rewriting skill specifically adapted for Gemini 3, Google’s reasoning-focused language model. It restructures general prompts for tools such as classifiers, reviewers, analysers, and agents.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A tool that restructures general AI prompts for classifiers, review tools, analyzers, and agents using eleven strengthening techniques. It is intended for prompts used with language models of different sizes.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A skill that evaluates how much effort an AI character’s design demands from the person role-playing with it. It scores 24 items across six types of burden and considers how that burden may grow during long conversations.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A diagnostic tool for AI character prompts, character settings, and conversation histories. It scores vulnerability to six types of failures that develop through a chain of seemingly normal conversation steps, and can estimate the current stage of a chain.

gpsnmeajp/ai-character-checker

Skill Claude CodeCodex

A Japanese-language skill for designing system instructions that define a consistent AI character's personality and behavior.