A structured framework to decompose technical designs into granular, 2-5 minute tasks with mandatory verification steps.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install writing-plans-2
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install writing-plans-2 using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
The Writing Plans skill is a robust methodology designed to bridge the gap between high-level architectural design and the actual implementation phase. By leveraging this framework within Openclaw Skills, developers and AI agents can transform abstract approvals into a concrete, ordered sequence of micro-tasks. This approach minimizes technical debt and implementation errors by ensuring every single change is small, focused, and immediately verifiable.
This skill emphasizes precision over ambiguity. Each plan generated provides a clear roadmap that includes exact file paths, implementation snippets, and the specific commands needed to verify success. By adopting this workflow, teams can maintain a high velocity while ensuring that every step of the development lifecycle is documented and testable.
To start using this skill with your Openclaw Skills setup, ensure your project structure supports a documentation directory for plans.
mkdir -p docs/plans
You can then instruct your AI agent to follow the Writing Plans template when moving from the design phase to the implementation phase.
The Writing Plans skill generates structured Markdown files stored in the project repository. Each file follows a specific taxonomy to ensure clarity for both humans and agents.
| Metadata Field | Description |
|---|---|
| Goal | A concise, one-sentence objective for the task. |
| Files | A list of exact file paths to be created or modified. |
| Implementation | Code snippets or pseudo-code representing the changes. |
| Verification | The bash command to run and the expected output to confirm success. |
| Estimated Time | The predicted duration of the task, ideally between 2-5 minutes. |
Files are organized by date and feature name: docs/plans/YYYY-MM-DD-feature-name.md.
Loading
A Socratic design refinement skill that ensures clear technical specifications through collaborative exploration and documentation before implementation.

An autonomous payment protocol enabling AI agents to conduct machine-to-machine micropayments using HTTP 402 and stablecoins.

A versatile social media automation engine for scheduling multi-platform content and managing audience engagement.

A maintenance utility that automatically detects, unlocks, and repairs stale or corrupted session files to prevent lock timeout errors.

A powerful AI agent skill to log meals, track nutrition progress, and manage calorie goals hands-free using Openclaw Skills.

A hyper-local weather monitoring tool that alerts you only when severe weather polygons physically intersect your specific address coordinates.








































