ai-pattern-detection

The AI Pattern Detection skill is a specialized utility designed for writers, editors, and documentation teams who rely on LLM-assisted workflows but require human-grade authenticity. It operates by scanning text for high-frequency algorithmic signatures often associated with automated generation, such as overused corporate buzzwords, formulaic transitions like 'Moreover' or 'Furthermore', and repetitive structural patterns. By flagging these markers, the skill ensures that technical documentation, marketing copy, and professional communications maintain a distinct, authentic voice that resonates with human readers.

• Automatically identifies critical patterns including vague intensifiers, academic passive voice, and performance-driven cliches.

• Provides actionable replacement suggestions to transform robotic, generic text into clear, direct, and specific prose.

• Evaluates text for structural variety, checking for excessive em-dash usage and repetitive list structures that are common LLM artifacts.

• Integrates with AIWG frameworks, including command-line access via the /writing-validator tool for deep analysis of content drafts.

• Encourages the inclusion of authenticity markers such as real-world trade-offs, domain-specific vocabulary, and specific technical opinions.

• This skill functions best when paired with the @AIWG_ROOT/agentic/code/addons/voice-framework/ for target style profiles.

• The agent processes content through a multi-stage validation pipeline: scanning for banned patterns, counting contextual frequency, and producing a 0-100 authenticity score.

• Users should rely on the suggested replacements for specific metrics (e.g., replacing 'dramatically improves' with actual performance statistics like 'reduces latency by 40%').

• Ideal for technical writers ensuring their documentation meets high standards of readability, clarity, and precision.

• Compatible with any prose or content generation task, triggering automatically when the user mentions writing quality, authenticity, or natural voice.