domain-planning

Decomposes feature requirements into parallel domain execution plans and merges them into a unified implementation plan.

Updated May 11, 2026
One-click install
npx skills add https://github.com/thachrocky12345/local-agent-train-workstation --skill domain-planning-thachrocky12345
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: domain-planning
Source: https://github.com/thachrocky12345/local-agent-train-workstation/tree/main/.claude/skills/domain-planning
Command: npx skills add https://github.com/thachrocky12345/local-agent-train-workstation --skill domain-planning-thachrocky12345

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Planning a feature that spans many backend apps and frontend modules often produces fragmented plans with conflicting assumptions about shared interfaces. This Skill fans a requirements document or Jira ticket out to per-domain planners and reconciles their outputs into one conflict-free, ordered execution plan. ## Core Features & Use Cases - Domain Decomposition: Maps requirements to a registry of domains (booking, payments, attribution, auth, etc.) with a hard cap of 8, producing a brief per domain with input/output contracts. - Parallel Domain Planning: Launches one planner agent per domain, each producing an execution plan with file-level changes, risk ratings, migrations, tests, and explicit assumptions about other domains. - Merge Reconciliation: Builds an integration matrix across domains, labels conflicts (C-01, C-02...), resolves them, and emits a unified plan with cross-domain execution ordering and an integration test checklist. - Use Case: Given a Jira ticket like RGDEV-205 for checkout discounts, the pipeline identifies Booking, Attribution, and Payments domains, plans each in parallel, resolves the discount-field ownership conflict, and outputs a sequenced unified plan. ## Quick Start Ask the assistant to run /domain-planning on a Jira ticket key, a requirements document path, or a plain feature description to generate a unified multi-domain implementation plan.

Frequently Asked Questions about domain-planning

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

FAQPage Schema
How do I plan a feature that spans multiple backend and frontend domains?

Run the domain-planning pipeline with a Jira ticket, requirements file, or plain description. It decomposes the feature into up to 8 domains, plans each in parallel with explicit interface contracts, then merges everything into a unified, conflict-resolved execution plan.

How does multi-agent planning detect conflicts between domains?

Each domain planner declares explicit assumptions and interface contracts it commits to. The merge reconciler builds an integration matrix comparing these contracts, flags mismatches as labeled conflicts (C-01, C-02), and records a resolution with the owning domain.

What inputs does the domain planning pipeline accept?

It accepts a Jira issue key fetched via the Atlassian MCP integration, a path to a requirements markdown file, or a plain-text feature description passed directly as the argument.

Is there a limit on how many domains a feature plan can cover?

Yes, there is a hard cap of 8 domains per plan. If decomposition identifies more than 8 touched domains, the smallest ones are consolidated before planners launch.

What happens after the unified plan is generated?

The unified plan is saved to /tmp/domain-plan-{slug}-unified.md with resolved conflicts, cross-domain execution ordering, and an integration test checklist. You can then execute it with an implement-from-plan step or run a scope audit per domain first.