| name | plan |
| description | Decompose a feature into atomic, verifiable implementation steps before coding. Use when the user describes a feature that spans multiple files or sessions, or asks for a roadmap, breakdown, or plan. |
| argument-hint | ["feature-description"] |
| allowed-tools | Read, Grep, Glob |
I am STARTING A PLANNING SESSION for the following feature:
$ARGUMENTS
Ref: .dev-context/claude-context.md for build constraints, architecture, coding conventions, and the Hard Rules.
PLANNING CONSTRAINTS
- Do not write implementation code. Focus only on the roadmap.
- Assume Seastar constraints apply. Flag any step that involves cross-shard communication or async complexity.
- Read before planning. Read every file the plan will touch. A plan that names the wrong file or a nonexistent function is worse than no plan.
TASK: Decompose the Feature
1. Break into Atomic Steps
Each step should be:
- Independent (can be implemented/tested in isolation)
- Small (completable in one focused session)
- Verifiable (has clear "done" criteria)
2. Identify Affected Files
For each step, list the files that will be modified:
Step 1: [description]
- src/foo.cpp
- src/foo.hpp
- tests/unit/foo_test.cpp
3. Flag Architectural Concerns
For each step, check:
4. Identify Dependencies
Which steps depend on others? Create a dependency graph if complex.
OUTPUT FORMAT
## Feature: [Name]
### Prerequisites
- [ ] Any setup or refactoring needed first
### Implementation Steps
#### Step 1: [Title]
- **Files:** src/x.cpp, src/x.hpp
- **Description:** [What this step accomplishes]
- **Concerns:** [Any Hard Rules that apply]
- **Status:** [ ] Pending
#### Step 2: [Title]
- **Depends on:** Step 1
- **Files:** ...
...
### Post-Implementation
- [ ] Run `/review` (Hard Rules compliance)
- [ ] Run `/doc` (tests, docs, BACKLOG)
- [ ] Deferred Gates for the developer's Docker build (exact `make` commands — see `/validate`)
Save output to BACKLOG.md under a new feature section. Then implement step-by-step with /implement.