kimchi:plan-synthesize

Merge cross-model plan revisions into a final executable plan.

8|Updated Jan 31, 2026
One-click install
npx skills add https://github.com/Tromml/kimchi --skill kimchi-plan-synthesize
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kimchi:plan-synthesize
Source: https://github.com/Tromml/kimchi/tree/main/plugins/kimchi/skills/plan-synthesize
Command: npx skills add https://github.com/Tromml/kimchi --skill kimchi-plan-synthesize

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill consolidates feedback from multiple AI models on a draft plan, resolving disagreements and integrating the best suggestions to create a superior, final plan.

Core Features & Use Cases

  • Cross-Model Analysis: Compares plan revisions from different AI models to identify consensus and unique insights.
  • Conflict Resolution: Artfully blends differing approaches from various models into a coherent, superior plan.
  • Use Case: After running a plan through Claude, Codex, and Gemini, use this Skill to merge their feedback, ensuring the final plan incorporates the strongest suggestions from all models, leading to a more robust and well-vetted task specification.

Quick Start

Synthesize the revised plans for the 'Add user authentication' feature.

Frequently Asked Questions about kimchi:plan-synthesize

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

FAQPage Schema
How do I merge AI plan revisions from multiple models into one final plan?

You can synthesize AI plan revisions by reading PLAN-REVISED-*.md, PLAN-DRAFT.md, and REQUIREMENTS.md files. The Skill merges cross-model feedback, resolves disagreements, and integrates the best suggestions to generate a final PLAN-SYNTHESIZED.md file.

What is cross-model plan synthesis for code generation?

Cross-model plan synthesis analyzes draft plan revisions from various AI models to identify consensus and unique insights. It artfully blends differing approaches into a coherent, superior plan by resolving conflicts and integrating the strongest suggestions.

How do I resolve disagreements between different AI models on a draft plan?

Resolving disagreements between AI models requires cross-model analysis that compares plan revisions to identify consensus and unique insights. The synthesis process blends differing approaches into a coherent, superior plan by integrating the strongest suggestions.

Do I need specific markdown files to synthesize a hybrid AI plan?

Yes, synthesizing a hybrid AI plan requires reading PLAN-REVISED-*.md, PLAN-DRAFT.md, and REQUIREMENTS.md files. These inputs provide the draft plan, cross-model revisions, and project requirements needed to generate the final PLAN-SYNTHESIZED.md output.

Can I use this plan synthesis approach after running drafts through Claude, Codex, and Gemini?

Yes, after running a plan through Claude, Codex, and Gemini, you can use this synthesis approach to merge their feedback. It ensures the final plan incorporates the strongest suggestions from all models, leading to a robust task specification.

What is the best way to consolidate multi-agent feedback into an executable plan?

The best way to consolidate multi-agent feedback is cross-model analysis that identifies consensus and unique insights. It blends differing approaches from various models into a coherent, superior plan to produce a final, executable task specification.