PRD to Dependency-Linked GitHub Issues
Turn a PRD into testable vertical-slice GitHub issues with acceptance criteria, verified dependencies, and parallel work lanes.
Before & after
Same prompt, same agent — the only difference is the skill.
Breaking down a roadmap into issues is smart for parallel work. I'd need to see the roadmap and understand your development flow—how do you define vertical slices? Are you aiming for tracer-bullet architectures that touch every layer, or smaller incremental steps? How granular should each issue be, and do you have a preference for which parts to tackle first? I can help you draft issues with blocking relationships, but knowing your team's velocity and testing conventions would help prioritize them.
VERTICAL SLICE BREAKDOWN: ROADMAP → ISSUES PARENT EPIC Title → "Build Real-time Collaboration" (PRD link) Status → Ready for slicing Estimated total → 40-50 hours across 5 slices VERTICAL SLICE 1: FOUNDATION Title → "Event stream schema and API contract" Type → Tracer bullet (touches schema, API, client) Blocked by → Nothing Blocks → Slices 2, 3, 4, 5 Estimate → 8 hours (design 3h, implement 4h, test 1h) Owner → Backend lead Checklist → Schema designed, API endpoint live, docs written Risk → Schema changes downstream (mitigate with contract tests) VERTICAL SLICE 2: WEBSOCKET PLUMBING Title → "Upgrade to WebSocket from polling" Blocked by → Slice 1 (API contract) Blocks → Slices 4, 5 Estimate → 6 hours Owner → Backend + DevOps Testing → Load test (100 concurrent), no regression on HTTP routes VERTICAL SLICE 3: CLIENT SYNC STATE Title → "Sync local state with server events" Blocked by → Slice 1 (event schema) Blocks → Slice 5 (UI) Estimate → 7 hours Owner → Frontend lead Testing → 10 edge cases (reorder, delete during sync, offline) VERTICAL SLICE 4: REAL-TIME UI UPDATES Title → "Render state changes as server events arrive" Blocked by → Slices 2, 3 Blocks → Nothing Estimate → 9 hours Owner → Frontend Testing → Concurrent edits, no race conditions VERTICAL SLICE 5: E2E FLOW & INTEGRATION Title → "End-to-end collaboration scenario (2 users editing)" Blocked by → Slices 1, 3, 4 Blocks → Nothing (release gate) Estimate → 8 hours Owner → QA + Full-stack review Testing → Record video of 2-user scenario, edge case checklist DEPENDENCY GRAPH Parallel work possible → Slices 2 & 3 can start after Slice 1 complete Critical path → 1 → 2 → 4 → 5 (30 hours) Non-blocking → Slice 3 can finish independently SUGGESTED SPRINT ALLOCATION Week 1 → Slice 1 (foundation) Week 2 → Slices 2 & 3 (parallel) Week 3 → Slices 4 & 5 (sequential due to blocking)
About this skill
Your PRD is approved, but the work queue still says “build the backend,” “build the frontend,” and “add tests.” This skill turns that plan into small, testable issues that deliver behavior across the required layers, with explicit dependencies so engineers and agents know what they can start.
What it does
- Finds the smallest end-to-end tracer bullet that proves the feature's riskiest integration.
- Writes child issues with scoped behavior, acceptance checks, exclusions, and implementation boundaries.
- Separates true prerequisites from preferred ordering and shared-file conflicts.
- Builds an acyclic blocking graph and identifies immediately actionable issues and parallel lanes.
- Maps every PRD requirement to an issue or an explicit deferral so scope cannot disappear between tickets.
- Creates and reads back GitHub issues and native dependency links when authorized access supports them; otherwise returns ready-to-file drafts with the access gap stated.
What you get
A compact execution plan, usually 3-8 child issues, a requirement coverage map, and a dependency graph. Each issue references the parent PRD and includes enough context for an engineer or agent to implement and verify it without reconstructing the whole plan.
Best for
Engineering leads and teams turning approved PRDs into pullable GitHub queues, especially when multiple engineers or coding agents need to work at the same time.
SKILL.md preview
---
name: prd-to-dependency-linked-github-issues
description: Convert a PRD into implementation-ready vertical-slice GitHub issues by tracing requirements to observable behavior, testing dependency edges, and separating execution blockers from scheduling conflicts. Use when the user says "turn this PRD into issues", "split this spec into vertical slices", "make tickets our agents can pick up", "what can we build in parallel?", "this feature is stuck behind one huge ticket", "our tickets are all backend then frontend", or "add blocking relationships to these issues".
---
# PRD to Dependency-Linked GitHub Issues
… (sign up to view the full skill)More development skills
View all Development skills →Performance Bottleneck Profiling: Node, Python, Go
Classifies a slowdown, gives exact profiler commands, reads the output, and proves the fix with same-conditions p95/p99 numbers.
Feature Scaffolding with Working Tests
Turn a short feature spec into code, runnable tests, and integration wiring that match your existing project.
Feature Flag Rollouts and Kill Switches
Design staged flag rollouts, verify rollback paths, and audit flag debt with explicit defaults, gates, and retirement plans.
Microcontroller Firmware and RTOS Engineering
Build and debug MCU firmware with bounded ISR/RTOS timing, safe DMA ownership, and measured power budgets.