program-refining

Refine epic plans into dependency-aware child feature seeds with durable artifacts.

1|Updated Mar 7, 2026
One-click install
npx skills add https://github.com/WASD6570/crypto-monitor --skill program-refining
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: program-refining
Source: https://github.com/WASD6570/crypto-monitor/tree/main/.agents/skills/program-refining
Command: npx skills add https://github.com/WASD6570/crypto-monitor --skill program-refining

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Large, vague epics under plans/epics/ often block delivery because they mix independent work, hide dependencies, and duplicate existing efforts; this Skill turns those broad epics into bounded, dependency-aware child features that are safe to hand off for feature-level planning.

Core Features & Use Cases

  • Context-aware analysis: Reads the epic overview and only the materially relevant initiative and program docs to avoid unnecessary scope creep.
  • Dependency-aware decomposition: Identifies completed work, missing pieces, and natural decomposition seams and proposes the smallest independently reviewable child features.
  • Durable refinement artifacts: Writes concise artifacts (90-refinement-map.md, 91-child-plan-seeds.md, 92-refinement-handoff.md) into the epic folder to enable downstream agents to continue without re-reading the whole epic.
  • Handoff and gating: Recommends which child feature to plan next and which seeds are safe to plan in parallel while explicitly avoiding creating active plans unless asked.

Quick Start

Refine the epic plans/epics/market-data/ using its 00-overview.md into dependency-aware child features and write 90-refinement-map.md, 91-child-plan-seeds.md, and 92-refinement-handoff.md into that epic folder.

Frequently Asked Questions about program-refining

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

FAQPage Schema
How do I decompose large product epics into dependency-aware child features?

Epic decomposition analyzes broad epic overviews alongside initiative and program docs to identify dependencies and produce the smallest independently reviewable child features for safe handoff.

What is the best way to identify dependencies and completed work before feature planning?

Dependency analysis reads the epic overview and relevant initiative documents to identify missing pieces, completed work, and natural decomposition seams, outputting a refinement map and child-plan seeds.

How do I refine epic plans without creating active feature plans prematurely?

Epic refinement generates durable artifacts like refinement maps and handoff documents in the epic folder, explicitly avoiding the creation of active feature plans unless explicitly requested by the user.

Can I use initiative and program documents to constrain scope during epic decomposition?

Yes, context-aware analysis reads materially relevant initiative and program docs during epic decomposition to constrain scope and prevent unnecessary scope creep when refining child features.

How do I determine which child features are safe to plan in parallel after epic decomposition?

Refinement handoff artifacts recommend which child feature to plan next and identify which child-plan seeds are safe to plan in parallel based on the dependency-aware decomposition analysis.