SDD Writing Plans is a specialized AI agent skill that converts high-level design specifications into granular, validated, and executable technical task lists.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install sdd-writing-plans
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install sdd-writing-plans using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is SDD Writing Plans?
SDD Writing Plans is a critical utility within the Openclaw Skills ecosystem designed to bridge the gap between software design and implementation. It parses spec-design.md files to generate a comprehensive spec-plan.md, which serves as a roadmap for AI coding agents. By focusing on principles like DRY (Don't Repeat Yourself) and YAGNI (You Ain't Gonna Need It), the skill ensures that development plans remain concise, intentional, and strictly focused on logic rather than boilerplate code.
This skill is engineered for precision, performing workspace configuration checks and model compatibility audits before execution. It automates the heavy lifting of breaking down a feature into discrete components, identifying impacted files, and establishing automated verification steps using CLI commands and scripts. This structured approach reduces ambiguity and ensures that every task has a clear path to successful acceptance.
SDD Writing Plans Use Cases
- Translating complex software architecture designs into actionable developer tasks.
- Generating automated testing and verification checklists for new features.
- Creating structured roadmaps for multi-step coding assignments in AI-driven workflows.
- Standardizing the execution process across different development teams using Openclaw Skills.
- Automating the identification of affected files and technical dependencies in a project workspace.
How SDD Writing Plans Works
- The skill initializes by reading the
.sdd-workspaceconfiguration to establish the root directory and ensure the environment is ready. - It performs a model check to verify if the agent is running on Opus for optimal reasoning; if not, it provides a non-blocking suggestion to switch.
- The agent locates the primary
spec-design.mdfile, either via a user-provided path or by scanning the workspace for the most recent design documents. - It explores the project context by scanning source files and existing patterns to inform the technical logic of the plan.
- A comprehensive list of tasks is generated, including file-specific implementation steps and automated verification commands (curl, grep, or scripts).
- The skill presents a final summary and task list to the user for review and confirmation.
- Upon approval, the
spec-plan.mdis saved to the relevant directory, ready for the execution phase.
SDD Writing Plans Setup
To use this skill within your workspace, ensure you have initialized your Openclaw environment:
# Initialize the global workspace if you haven't already
/sdd-global-init
# Run the writing plans skill on a specific design doc
/sdd-writing-plans {workspace}/spec/feature_name/spec-design.md
Note: A valid .sdd-workspace file must exist in the current directory for the skill to function correctly.
SDD Writing Plans Data Schema & Taxonomy
The skill follows a strict directory and file taxonomy to maintain project organization:
| File | Location | Purpose |
|---|---|---|
.sdd-workspace |
Root | Stores the workspace_path and configuration. |
spec-design.md |
{workspace}/spec/{feature}/ |
The source design document used as input. |
spec-plan.md |
{workspace}/spec/{feature}/ |
The generated execution plan containing tasks and checklists. |
Task Schema
Each task in the generated plan includes:
- Involved Files: List of new and modified file paths.
- Execution Steps: Intent-based descriptions with checkbox formatting.
- Verification Steps: CLI-executable commands and their expected outputs.
SDD Writing Plans Advanced Features
- Multi-step Feature Discovery: Automatically identifies and ranks design documents by modification time for quick selection.
- Automated Model Auditing: Detects agent model capabilities and recommends high-reasoning models for complex planning.
- End-to-End Acceptance Mapping: Dynamically generates a final acceptance task with scenarios derived directly from the design spec.
- Non-Blocking Recommendations: Provides technical guidance without interrupting the automated workflow.
- Integrated Workspace Validation: Ensures environment consistency before any file operations are performed.
SKILL.md
Loading
Related Openclaw Skills
Ecommerce Copy Humanizer ID
A specialized tool to transform robotic or translated Bahasa Indonesia ecommerce text into natural, conversion-ready copy.
Formula Decoder AI
An expert AI persona that transforms abstract mathematical and physical formulas into intuitive, visual, and real-world mental models.
SnapAPI
A comprehensive web intelligence API that enables AI agents to screenshot, extract structured data, and monitor websites for changes.
TicketMaster
A specialized AI agent skill for searching Ticketmaster events, venues, and attractions using the Discovery API with market-aware filters.
DeFi Consensus Oracle (SuperColony)
A decentralized swarm intelligence oracle that aggregates independent AI agent opinions into cryptographically verified market consensus.
Cursor IDE AI Expert
A comprehensive configuration expert for mastering Cursor IDE AI features, rule-based automation, and Agent mode workflows.
Featured*
