Refactor high-complexity React components in Dify frontend. Use when `pnpm analyze-component...
npx skills add miles-knowbl/orchestrator --skill "roadmap-tracker"
Install specific skill from multi-skill repository
# Description
Maintain and evolve project roadmaps based on execution history, learnings, and re-prioritization signals.
# SKILL.md
name: roadmap-tracker
description: "Maintain and evolve project roadmaps based on execution history, learnings, and re-prioritization signals."
phase: COMPLETE
category: meta
version: "1.0.0"
depends_on: ["retrospective"]
tags: [meta, planning, roadmap, milestones, prioritization]
Roadmap Tracker
Maintain project roadmaps that evolve from execution learnings.
When to Use
- After loop completion β Update roadmap with what was learned
- Planning sessions β Review and re-prioritize the roadmap
- Milestone tracking β Check progress against planned milestones
- When you say: "update roadmap", "what's next", "milestone status"
Required Deliverables
| Deliverable | Location | Condition |
|---|---|---|
| Roadmap file | ROADMAP.md or memory/roadmap.json |
Always |
Core Concept
Roadmap tracking answers: "What's planned, what's done, and what should change based on what we learned?"
Retrospective Insights β Roadmap Update β Re-prioritization β Next Loop Planning
Roadmaps are living documents that evolve with every loop iteration.
Roadmap Structure
# Project Roadmap
## Current Milestone: v1.0 Launch
- [x] Authentication system
- [x] Core API endpoints
- [ ] Admin dashboard
- [ ] Email notifications
## Next: v1.1 Polish
- [ ] Performance optimization
- [ ] Accessibility audit
- [ ] Documentation site
## Backlog
- Rate limiting
- Webhook system
- Multi-tenancy
Milestone Schema
{
"id": "v1.0",
"title": "v1.0 Launch",
"status": "in-progress",
"items": [
{ "id": "auth", "title": "Authentication", "status": "done", "loopId": "exec-001" },
{ "id": "api", "title": "Core API", "status": "done", "loopId": "exec-002" },
{ "id": "admin", "title": "Admin dashboard", "status": "planned" }
],
"confidence": 0.8,
"updatedAt": "2026-01-26"
}
Re-prioritization Triggers
| Trigger | Action |
|---|---|
| Retrospective reveals blocker | Move blocker to current milestone |
| Feature took longer than expected | Re-estimate remaining items |
| User feedback changes priorities | Reorder backlog |
| Technical discovery | Add new items or remove obsolete ones |
| Dependency change | Reorder based on new constraints |
Checklist
- [ ] Roadmap exists and is current
- [ ] Completed items linked to execution IDs
- [ ] Next milestone clearly defined
- [ ] Backlog prioritized
- [ ] Confidence score reflects reality
- [ ] Updated after every loop completion
Relationship to Other Skills
| Skill | Relationship |
|---|---|
retrospective |
Retro insights trigger roadmap updates |
priority-matrix |
Prioritization feeds into roadmap ordering |
requirements |
Requirements flow from roadmap items |
estimation |
Estimates inform milestone confidence |
Key Principles
Living document. Roadmaps change with every iteration β that's a feature, not a bug.
Link to executions. Every completed item should reference the loop that delivered it.
Confidence over dates. Express milestone likelihood, not fixed deadlines.
Backlog is a queue. Items move forward based on priority, not wishful planning.
# Supported AI Coding Agents
This skill is compatible with the SKILL.md standard and works with all major AI coding agents:
Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.