Agent skill
plan
Guided implementation planning with codebase understanding and requirement focus
Install this agent skill to your Project
npx add-skill https://github.com/SmidigStorm/storm-claude-marketplace/tree/main/plugins/pm-planning/skills/plan
SKILL.md
Implementation Planning
You are helping a developer create an implementation plan for a backlog item. Follow a systematic approach: understand the requirements, explore the codebase with specialized agents, ask clarifying questions, then design the implementation plan together.
Core Principles
- Ask clarifying questions: Identify all ambiguities, edge cases, and underspecified behaviors. Ask specific, concrete questions rather than making assumptions. Wait for user answers before proceeding.
- Understand before planning: Read and comprehend existing code patterns first
- Read files identified by agents: When launching agents, they return lists of important files. Read those files to build detailed context.
- Interactive: Every major decision involves the user
- Traceable: Link plan steps to requirements
- Use TodoWrite: Track all progress throughout
Phase 1: Discovery
Goal: Understand what needs to be built
Initial request: $ARGUMENTS
Actions:
- Create todo list with all phases
- If backlog item unclear, ask user for:
- What problem are they solving?
- What should the feature do?
- Who benefits from this?
- Any constraints or requirements?
- Summarize understanding and confirm with user
Phase 2: Requirements Mapping
Goal: Identify all requirements this backlog item addresses
Actions:
- List any known requirements (ask user)
- For each requirement, clarify:
- What is the acceptance criteria?
- What are the business rules?
- Any edge cases or examples?
- Use DOM-SUB-CAP-NNN format if requirements exist in the system
- Confirm scope with user before proceeding
Phase 3: Codebase Exploration
Goal: Understand relevant existing code and patterns at both high and low levels
Actions:
-
Ask user: "Are there specific areas of the codebase I should look at?"
-
Launch 2-3
code-exploreragents in parallel. Each agent should:- Trace through the code comprehensively
- Focus on getting a comprehensive understanding of abstractions, architecture and flow of control
- Target a different aspect of the codebase
- Include a list of 5-10 key files to read
Example agent prompts:
- "Find features similar to [feature] and trace through their implementation comprehensively"
- "Map the architecture and abstractions for [feature area], tracing through the code comprehensively"
- "Analyze the current implementation of [existing feature/area], tracing through the code comprehensively"
-
Once the agents return, read all files identified by agents to build deep understanding
-
Present comprehensive summary of findings and patterns discovered
-
Ask user if findings match their expectations
Phase 4: Clarifying Questions
Goal: Fill in gaps and resolve all ambiguities before designing
CRITICAL: This is one of the most important phases. DO NOT SKIP.
Actions:
- Review requirements and codebase findings
- Identify underspecified aspects:
- Edge cases and error handling
- Integration points
- Scope boundaries
- Performance needs
- Backward compatibility
- Present all questions to the user in a clear, organized list
- Wait for answers before proceeding to plan design
If the user says "whatever you think is best", provide your recommendation and get explicit confirmation.
Phase 5: Plan Design
Goal: Design the implementation approach collaboratively using architecture expertise
Actions:
-
Launch 2-3
code-architectagents in parallel with different focuses:- Minimal changes: Smallest change, maximum reuse of existing code
- Clean architecture: Maintainability, elegant abstractions
- Pragmatic balance: Speed + quality
-
Review all approaches and form your opinion on which fits best for this task
- Consider: small fix vs large feature, urgency, complexity, team context
-
Present to user:
- Brief summary of each approach
- Trade-offs comparison
- Your recommendation with reasoning
- Concrete implementation differences
-
Ask user which approach they prefer
-
Break down chosen approach into concrete implementation steps
-
Map each step to requirements it implements (DOM-SUB-CAP-NNN)
-
Confirm plan with user before documenting
Phase 6: Document Plan
Goal: Create the plan document
Output: docs/plans/[backlog-item-name].md
Actions:
- Create plan document with:
- Backlog item summary
- Requirements in scope (with IDs)
- Codebase patterns to follow
- Implementation steps with file paths
- Acceptance criteria
- Open questions (if any)
- Decisions made (with rationale)
- Mark all todos complete
- Present summary to user
Plan Document Template
# [Backlog Item Title]
## Summary
[What this plan accomplishes]
## Requirements
- [ ] DOM-SUB-CAP-001: [Title] - [Acceptance criteria]
- [ ] DOM-SUB-CAP-002: [Title] - [Acceptance criteria]
## Architecture Approach
[Which approach was chosen and why]
## Codebase Patterns
- [Pattern 1]: Found in `path/to/file.ts:line`
- [Pattern 2]: Found in `path/to/file.ts:line`
## Implementation Steps
### Step 1: [Description]
**Implements**: DOM-SUB-CAP-001
**Files**:
- `path/to/file.ts` - [What to change]
### Step 2: [Description]
**Implements**: DOM-SUB-CAP-001, DOM-SUB-CAP-002
**Files**:
- `path/to/new-file.ts` - [Create new file for X]
- `path/to/existing.ts` - [Modify Y]
## Acceptance Criteria
- [ ] [Criterion from requirements]
- [ ] [Criterion from requirements]
## Open Questions
- [Any unresolved items to address during implementation]
## Decisions Made
- [Decision 1]: [Rationale]
- [Decision 2]: [Rationale]
Resources
- For planning dialogue patterns and checklists, see planning-guide.md
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
architecture
Manage tech stack and architecture decisions
openclaw
This skill should be used when the user asks to "install openclaw", "set up openclaw", "configure openclaw channels", "openclaw gateway", "openclaw troubleshooting", "add whatsapp to openclaw", "add telegram to openclaw", "openclaw agent", "openclaw models", "openclaw cron", "openclaw memory", "openclaw skills", "openclaw security", "openclaw browser", "openclaw sessions", "openclaw workspace", "openclaw docker", "openclaw not working", "debug openclaw", "connect discord to openclaw", "openclaw setup guide", or mentions openclaw setup, configuration, troubleshooting, or daily usage.
vision
Create or update the product vision document
requirements
Write BDD requirements in Gherkin format. Guides the user through the process.
domain
Document domain knowledge including entities, processes, and ubiquitous language
execute
Execute an implementation plan step by step
Didn't find tool you were looking for?