Development planning and architecture
Scanned 9/2/2026
Install to Claude Code
npx -y skills add Smart-AI-Memory/attune-ai --skill plan --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Plan?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/smart-ai-memory-plan)More formats (shields.io, HTML) on the badges page.
---
name: plan
description: Development planning and architecture
---
# plan
Development planning and architecture design.
## Routes
| Subcommand | Action |
| ---------- | ------ |
| `feature` | Plan a new feature |
| `refactor` | Plan refactoring |
| `architecture` | Architecture review |
## Usage
```bash
/plan # Ask what to plan
/plan feature # Plan a feature
/plan refactor # Plan refactoring
/plan architecture # Architecture review
```
## Behavior
All planning tasks use `EnterPlanMode` to create
a structured plan for user approval before any
implementation.
**After every plan is approved**, follow the
[Post-Plan Handoff](#post-plan-handoff) process.
### feature
Use `AskUserQuestion` to understand:
- What feature? What problem does it solve?
- What's the scope? Which files/modules?
- Any constraints or preferences?
Then use `EnterPlanMode` to design the implementation
plan. After approval, follow Post-Plan Handoff.
### refactor
Use `AskUserQuestion` to understand:
- What code needs refactoring?
- What's the goal? (simplify, split, extract)
- Any constraints?
Then use `EnterPlanMode` to plan the refactoring.
After approval, follow Post-Plan Handoff.
### architecture
Use `AskUserQuestion` to understand:
- What system or subsystem to review?
- Any specific concerns?
Then analyze the codebase structure and provide
architectural recommendations. If the review
produces actionable changes, follow Post-Plan
Handoff.
## Post-Plan Handoff
**This section applies to ALL routes after a plan
is approved.**
### Step 1: Save the plan
Save the approved plan to `.claude/plans/` using
the [Plan File Format](#plan-file-format) below.
File naming convention:
`{route}-{slug}-{YYYY-MM-DD}.md`
Examples:
- `refactor-auth-module-2026-02-22.md`
- `feature-dark-mode-2026-02-22.md`
Ensure `.claude/plans/` directory exists before
writing.
### Step 2: Offer execution
Use `AskUserQuestion` to ask:
```yaml
question: "Plan saved to .claude/plans/{filename}.
Ready to execute?"
header: "Next"
options:
- label: "Execute now"
description: "Start implementing — I'll carry
the plan context forward so you won't need
to re-explain anything"
- label: "Save for later"
description: "Plan is saved — pick it up later
with /dev {route}"
```
### Step 3: Transition
- **Execute now**: Transition into the corresponding
route with full plan context. Do NOT re-ask
scoping questions — the plan already contains
all needed context. Use this mapping:
| Plan Route | Execute Via |
| ---------- | ----------- |
| `/plan feature` | `/dev` (implement directly) |
| `/plan refactor` | `/dev refactor` |
| `/plan architecture` | (analysis only — no auto-execute) |
- **Save for later**: Confirm the file path and
end gracefully. Tell the user they can resume
with `/dev {route}` in a future session.
## Plan File Format
All saved plans MUST use this format:
```markdown
# {Title}
**Created:** {YYYY-MM-DD}
**Source:** /plan {route}
**Route:** {route}
**Status:** pending
## Problem
{1-2 sentence problem statement}
## Goals
- {Must-have 1}
- {Must-have 2}
- {Nice-to-have (marked as such)}
## End State
{Concrete description of what "done" looks like}
## Scope
- **Files:** {list of target files/directories}
- **Type:** {refactor | feature | architecture}
## Approach
1. {Step 1 with specific file references}
2. {Step 2}
3. {Step 3}
## Open Questions
- {Anything unresolved}
```
The **Route** and **Status** fields are required
for `/dev` plan detection. Update **Status** to
`in-progress` when execution begins and `completed`
when done.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!