A meta-cognitive engine that forces AI agents to think, retrieve memories, and plan optimal paths before taking action.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install thinking-skill
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 thinking-skill 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 Thinking Skill - Deep Thought and Planning?
The Thinking Skill serves as a sophisticated meta-cognitive framework designed to elevate AI agent performance. By implementing a systematic six-step workflow—task analysis, skill inventory, memory retrieval, path planning, risk assessment, and execution—it ensures that agents do not act blindly. Instead, they leverage existing experiences and evaluate technical constraints to deliver high-quality outcomes.
This skill is particularly valuable for complex, multi-step workflows where coordination between different tools is required. By integrating Openclaw Skills into the planning phase, the agent can identify the most efficient route, anticipate potential risks like API rate limits, and align its output with specific user preferences derived from historical data.
Thinking Skill - Deep Thought and Planning Use Cases
- Handling complex multi-step tasks requiring multiple skill coordination.
- Navigating unfamiliar domains by retrieving relevant background knowledge.
- Making critical decisions that require weighing pros and cons.
- Solving repetitive problems by searching for verified historical solutions.
How Thinking Skill - Deep Thought and Planning Works
- Task Analysis: Identify task types, key elements, and implicit user needs to define success criteria.
- Skill Inventory: Match available tools and capabilities to the specific requirements of the task.
- Memory Retrieval: Search past experiences and domain knowledge to find successful patterns or pitfalls to avoid.
- Path Planning: Design the optimal execution plan based on speed, success rate, and resource cost.
- Risk Assessment: Anticipate potential failures like timeouts or permission issues and prepare mitigation strategies.
- Start Execution: Begin the action sequence only after the plan is verified and risks are managed.
Thinking Skill - Deep Thought and Planning Setup
To activate the deep thinking mode, use the specific trigger command within your agent environment. Use the following command to enter the thinking state:
/thinking
Once activated, describe your complex task. Alternatively, use natural language triggers such as "think about this task" to initiate the Openclaw Skills thinking workflow automatically.
Thinking Skill - Deep Thought and Planning Data Schema & Taxonomy
The skill manages its state and organizational structure through a defined TypeScript interface and systematic Markdown templates:
| Component | Description |
|---|---|
| ThinkingState | Tracks activation status, timestamps, and timeouts (default 5 mins). |
| Task Analysis Template | Structured Markdown for identifying constraints and implicit needs. |
| Path Planning Table | Breaks down steps, tools used, and estimated time for completion. |
| Risk Matrix | Categorizes probability and mitigation for identified technical risks. |
Thinking Skill - Deep Thought and Planning Advanced Features
- Multi-agent coordination support through structured execution paths.
- Automated timeout management to prevent agents from getting stuck in loops.
- Memory search integration using vector or keyword retrieval from historical sessions.
- Dynamic fallback planning if the primary execution path encounters a risk trigger.
SKILL.md
Loading
Related Openclaw Skills
Open Poker Bot Builder
A comprehensive development toolkit for building, deploying, and optimizing autonomous AI poker bots on the competitive Open Poker platform.
Feishu Voice Interaction
An AI agent skill that enables natural voice-to-voice communication within Feishu using advanced text-to-speech models.
Temp Heartbeat Task Manager
A flexible task manager for AI agents that creates one-time scheduled heartbeats which automatically self-destruct after execution.
Skill Setup Flow
A standardized automation utility for initializing and configuring new Openclaw Skills within an AI agent ecosystem.
LMail Ops Complete
A comprehensive automation suite for LMail that handles everything from secure agent registration and PoW challenges to threaded inbox management.
img-ocr
A Tesseract-powered OCR utility designed to extract multilingual text from images and screenshots within an AI agent workflow.
Featured*
