By name: llm-writing type: guardrail description: Use when producing written artifacts for humans. Recognize and replace default LLM writing patterns. model-invocable: true
LLM Writing
Load /intent-modeling if it isn't already loaded.
Default LLM writing has recognizable patterns. Spot them, then decide whether each serves your purpose:
Behavioral Pulls
Filling structure without anchoring to purpose. Summarizing with labels instead of explaining how things work. Stating conclusions without evidence. Smoothing over uncertainty. Encoding corrections as prohibitions. Defining by negation. Restating what was just said as a transition.
Not always wrong; the failure is when they happen by default.
Conversational Bleed
Writing as if responding to a user when producing a document for a reader who wasn't in the conversation. "It's not X — it's Y" corrects a misconception nobody has. "Let's break this down" addresses a question nobody asked.
Write for the reader. They have no context from the conversation that prompted the document.