interleave

Generate documents by combining deterministic templates with LLM-driven content insertion.

3|1|Updated Feb 15, 2026
One-click install
npx skills add https://github.com/mistakeknot/Demarch --skill interleave
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: interleave
Source: https://github.com/mistakeknot/Demarch/tree/main/.gemini/generated-skills/interleave
Command: npx skills add https://github.com/mistakeknot/Demarch --skill interleave

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the challenge of generating documents efficiently by combining deterministic templating with targeted Large Language Model (LLM) interventions, optimizing token usage.

Core Features & Use Cases

  • Token-Efficient Generation: Reduces LLM costs and processing time by using templates for boilerplate content and LLMs only for dynamic sections.
  • Structured Document Creation: Facilitates the creation of complex documents where parts are fixed and others require creative or data-driven input.
  • Use Case: Generating personalized reports where standard sections are templated, but specific analysis or recommendations are dynamically produced by an LLM.

Quick Start

Activate the interleave skill to generate a document using a predefined template and LLM-driven content.

Frequently Asked Questions about interleave

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

FAQPage Schema
How do I reduce token usage when generating structured documents with dynamic content?

Generate documents efficiently by applying deterministic templates for fixed boilerplate and using LLM interventions only for dynamic sections. This hybrid approach to document generation ensures token efficiency by minimizing LLM processing to targeted content insertion, reducing overall costs.

What is the best way to create personalized reports that combine fixed layouts with dynamic analysis?

Create personalized reports by combining deterministic templating for standard sections with LLM-driven content insertion for dynamic analysis. This approach ensures structured document creation where specific recommendations are dynamically produced while maintaining fixed layouts for boilerplate content.

How does LLM-driven content insertion work with deterministic templates?

LLM-driven content insertion works by populating specific dynamic sections of a deterministic template. Fixed boilerplate content is handled by the template, while the LLM generates only the creative or data-driven elements required, resulting in token-efficient document generation.

Can I use this approach for complex form generation requiring data-driven input?

Yes, you can use this hybrid approach for complex form generation. It facilitates structured document creation where parts are fixed and others require data-driven input, making it applicable for scenarios needing dynamic elements like personalized reports.

Do I need specific project guidelines to use LLM islands for document generation?

Yes, using LLM islands for document generation requires careful alignment with established project philosophies and implementation guidelines. This ensures the templated structure and dynamic LLM content match the project's standards.