Stats
Actions
Tags
From decaf-planning
Turn a PRD into a multi-phase implementation plan using tracer-bullet vertical slices, then create work items (GitHub Issues, Azure DevOps, Beans) or save as a local Markdown file. Use when user wants to break down a PRD, create an implementation plan, plan phases from a PRD, or mentions "tracer bullets".
How this skill is triggered — by the user, by Claude, or both
Slash command
/decaf-planning:prd-to-planThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Break a PRD into a phased implementation plan using vertical slices (tracer bullets). Output work items to a tracking system or fall back to a Markdown file.
Break a PRD into a phased implementation plan using vertical slices (tracer bullets). Output work items to a tracking system or fall back to a Markdown file.
The PRD should already be in the conversation. If it isn't, ask the user to paste it or point you to the file.
If you have not already explored the codebase, do so to understand:
Check the project's CLAUDE.md for a "When Executing Plans" section (or similar deviation rules). If none exists, offer to add one. This defines what the AI can decide autonomously vs. what requires human input during plan execution.
## When Executing PlansWhen working from a plan or work item:
Present the template to the user and let them adjust the boundaries before adding it. If the user declines, proceed without it.
This step can run in parallel with steps 2 and 4.
@../../../../conventions/work-items.md
Detect the available system and confirm with the user. This can run in parallel with steps 2 and 3.
Before slicing, identify high-level decisions that are unlikely to change throughout implementation. Examples (adapt to the project type):
Express these using the project's language idioms (e.g., structs for Go, records/classes for C#, structs/enums for Rust).
These go in the root work item body (or plan header for markdown) so every phase can reference them.
Break the PRD into tracer bullet phases. Each phase is a thin vertical slice that cuts through ALL integration layers end-to-end, NOT a horizontal slice of one layer.
- Each slice delivers a narrow but COMPLETE path through every layer identified in the codebase exploration - A completed slice is demoable or verifiable on its own - Prefer many thin slices over few thick ones - Do NOT include specific file names, function names, or implementation details that are likely to change as later phases are built - DO include durable decisions: route paths, schema shapes, data model names - DO include scope boundaries per phase (see below)Scope boundaries per phase: Each phase should include a brief "Touches" and "Off limits" section. This prevents scope creep during implementation — Claude won't "helpfully" refactor adjacent code or add features that belong to a different phase.
Present the proposed breakdown as a numbered list. For each phase show:
Ask the user:
Iterate until the user approves the breakdown.
Create work items using the target system from step 4 and the conventions in work-items.md.
Root work item:
Plan: <Feature Name>One child work item per phase:
Phase N: <Title>For markdown output, use the template below instead.
# Plan:Source PRD:
Durable decisions that apply across all phases:
User stories:
A concise description of this vertical slice. Describe the end-to-end behavior, not layer-by-layer implementation.
Touches: <modules, layers, or subsystems this phase modifies> Off limits: <areas explicitly out of scope, even if related>
User stories:
...
Touches: ... Off limits: ...
npx claudepluginhub alphaleonis/decaf-claude-config --plugin decaf-planningSplits plans, specs, or PRDs into independent vertical-slice issues for an issue tracker. Use when converting a plan into actionable implementation tickets.
Turns requirements into structured implementation plans with subtasks, dependencies, file paths, and test scenarios. Use after brainstorming or to formalize clear requirements before coding.