kitt-create-meta-prompts

Automate multi-stage prompt workflows for Claude-to-Claude pipelines with XML metadata.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/utahdws/webteam-ai-tools --skill kitt-create-meta-prompts
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kitt-create-meta-prompts
Source: https://github.com/utahdws/webteam-ai-tools/tree/main/.ai-cli/shared/skills/kitt-create-meta-prompts
Command: npx skills add https://github.com/utahdws/webteam-ai-tools --skill kitt-create-meta-prompts

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill automates the creation and orchestration of multi-stage prompt workflows, enabling teams to generate, reference, and execute prompts for research, planning, and implementation with consistent provenance.

Core Features & Use Cases

  • Intake & topic identification: captures purpose and topic and initializes a new .prompts folder.
  • Chain detection & provenance: detects existing research/plan artifacts and references them.
  • Prompt generation & execution planning: builds purpose-specific prompts from patterns and intelligence rules.
  • Metadata & summaries: enforces XML metadata in outputs, generates SUMMARY.md, and supports versioning for changelog-like tracking.

Quick Start

Describe the task to capture purpose and topic, then let the system generate a new numbered prompt folder under .prompts with a SUMMARY.md.

Frequently Asked Questions about kitt-create-meta-prompts

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

FAQPage Schema
How do I automate multi-stage prompt workflows for Claude-to-Claude pipelines?

Automating multi-stage prompt workflows involves coordinating intake, chain detection, prompt generation, and execution planning across research, plan, execution, and refinement stages with structured XML metadata for reproducible outputs.

What is the best way to generate structured XML metadata for research and plan prompts?

Generating structured XML metadata for research and plan prompts is achieved by enforcing specific metadata formats in outputs, enabling automatic dependency tracking and versioned summaries for consistent provenance across prompt execution stages.

Can I track prompt provenance and dependencies across multiple execution stages?

Tracking prompt provenance and dependencies is supported through chain detection of existing research and plan artifacts, referencing them automatically, and generating SUMMARY.md files with versioning for changelog-like tracking of prompt workflows.

How do I organize and store generated prompts and artifacts for reproducible execution?

Organizing and storing generated prompts requires initializing a numbered prompt folder under a .prompts directory, storing artifacts there, and generating SUMMARY.md files with explicit artifact locations to support reproducible prompt workflows.

When should I use multi-stage prompt engineering instead of single prompt execution?

Multi-stage prompt engineering is needed when tasks require distinct research, planning, execution, and refinement phases with consistent provenance, automatic dependency tracking, and versioned summaries that single prompt execution cannot adequately provide.