section-writer

Writes a single document section from an outline node, brand voice spec, and KB context.

Updated May 23, 2026
One-click install
npx skills add https://github.com/TheGraphicKing/skott-canvas --skill section-writer-thegraphicking
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: section-writer
Source: https://github.com/TheGraphicKing/skott-canvas/tree/main/skills/section-writer
Command: npx skills add https://github.com/TheGraphicKing/skott-canvas --skill section-writer-thegraphicking

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Writing an entire document in one pass causes structural drift, uneven depth, and inconsistent tone across sections. This Skill writes one section at a time, bounded by an outline node, so each section can be reviewed, iterated, and approved independently while staying grounded in approved source material. ## Core Features & Use Cases - Single-Section Writing: Produces publish-ready prose for one H2 or H3 outline node without overstepping into adjacent sections. - Brand Voice Enforcement: Applies tone modifiers, banned vocabulary lists, and sentence structure patterns from a brand voice spec during generation, with a post-generation scan before delivery. - KB-Grounded Content: Integrates only facts and evidence from assigned knowledge base chunks, preventing invented statistics or unsupported claims. - Use Case: After approving a blog post outline, invoke this Skill for each section to generate a 600-word section that cites the assigned KB chunks, matches the brand voice, and lands within 15% of the target word count. ## Quick Start Write the 'Pricing Considerations' section from my approved outline using the brand voice spec and the three KB chunks assigned to that section.

Frequently Asked Questions about section-writer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write a document section by section with AI?▼

Provide an approved outline with H2 or H3 nodes, then invoke the section-writer once per node with its brief, word count target, and assigned KB chunks. Each section is generated and reviewed independently before moving to the next.

How to enforce brand voice in AI-generated content?▼

Pass a brand voice spec containing tone modifiers, banned vocabulary, and preferred sentence structures as an input. The Skill applies these constraints during generation and runs a post-generation scan for banned words and passive voice before delivery.

Can AI writing avoid making up facts and statistics?▼

Yes, by grounding generation in provided knowledge base chunks. This Skill only uses facts traceable to the assigned KB chunks or brand guidelines, and reports which citations were incorporated in its output.

What inputs does section-by-section content generation require?▼

It requires an outline node with header text, section brief, and word count target, plus a brand voice spec and KB context chunks. Optional inputs include a preceding section summary and funnel stage to adjust depth and assertiveness.

When should I not use single-section writing?▼

Avoid it for very short content under section scale, such as single paragraphs or social posts, where the outline and KB overhead adds no value. It is designed for sections of roughly 400 words or more within multi-section documents.