A methodology-driven skill that applies triage frameworks and cognitive load reduction strategies to any incoming communication stream.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install inbox
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 inbox 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 Inbox Skill?
The Inbox skill is a sophisticated decision-making framework designed to help AI agents manage the overwhelming influx of information across multiple platforms. Unlike standard plugins that simply fetch data, this skill provides the logical infrastructure to categorize, prioritize, and process messages from email, chat, social media, and project management tools. It focuses on the human element of productivity, ensuring that Openclaw Skills can help users stay focused by acting as an intelligent buffer between raw data and actionable tasks.
By implementing the Inbox skill, agents gain the ability to minimize context switching and reduce the anxiety associated with high unread message counts. It provides a standardized way to handle items based on their actual importance rather than their chronological order, making it an essential tool for developers building high-performance agentic workflows.
Inbox Skill Use Cases
- Triaging high-volume notification streams from Jira, GitHub, and Slack into actionable buckets.
- Reducing user anxiety by surfacing only critical items while hiding raw unread counts.
- Managing multi-channel communications where high-priority signals appear across different platforms.
- Matching daily inbox processing tasks to the user's current energy levels and time availability.
- Resurfacing aging tasks and identifying avoidance patterns in project communications.
How Inbox Skill Works
- The user provides a text-based representation of their current inbox or message stream to the agent.
- The skill analyzes the input and classifies each item into specific buckets: Requires decision, Requires awareness, Can be delegated, or Noise.
- The agent calculates the 'age' of items, flagging those that have been pending for 3 or more days to prevent them from slipping through the cracks.
- Items are batched by type or project to allow the user to process similar tasks together, significantly reducing cognitive load.
- Based on the user's specified time and energy state, the skill offers a curated selection of tasks, such as quick approvals or deep-work responses.
- The agent suggests the optimal response type, such as a pre-approved template, a holding response, or a full draft.
Inbox Skill Setup
To deploy this skill within your agentic environment, include the framework files in your skill configuration directory. Because this skill focuses on methodology rather than API integration, no external credentials are required.
# Navigate to your agent skills directory
cd ~/.openclaw/skills
# Clone or add the Inbox skill repository
git clone https://github.com/openclaw/skill-inbox.git inbox
Once added, ensure your agent is configured to reference the triage.md and responses.md files to utilize these Openclaw Skills effectively.
Inbox Skill Data Schema & Taxonomy
The Inbox skill utilizes a conceptual data schema to organize communication metadata and workflow states.
| Metadata Field | Type | Description |
|---|---|---|
| Bucket | Enum | Classification: Decision, Awareness, Delegate, Noise |
| Age Signal | String | Status: Pending (3d+), Concerning (7d+), or Buffer Calculation |
| Energy State | String | User capacity: Low Energy, Focused, or Need a Win |
| Response Type | String | Action: Template, Draft, Holding, or Full Compose |
| Channel Source | String | Origin: Email, Slack, Discord, Jira, etc. |
Inbox Skill Advanced Features
- Avoidance Pattern Detection: Automatically flags items that have been snoozed more than three times and suggests breaking them into smaller tasks.
- Multi-Channel Orchestration: Detects when the same contact is reaching out across multiple platforms and escalates the priority accordingly.
- Cognitive Load Filtering: Specifically designed to hide overwhelming 'unread' numbers and only show actionable item counts.
- Energy-Aware Task Delivery: Dynamically adjusts the complexity of tasks presented based on the user's available time and mental bandwidth.
SKILL.md
Loading
Related Openclaw Skills
Deep Research
A systematic multi-source investigation skill for AI agents that prioritizes methodology tracking and source evaluation over surface-level searches.
In-App Purchases
A comprehensive toolkit for implementing, managing, and optimizing in-app purchases and subscriptions across mobile platforms.
AI Image Generation
A professional-grade integration for generating and editing AI images across multiple providers using optimized prompt engineering and model selection.
AI Image Editing
A comprehensive AI-driven image manipulation skill for professional-grade inpainting, outpainting, background removal, and high-resolution upscaling.
Indie Hacker
A high-performance AI partner designed to help solo founders build, validate, and scale profitable software products with extreme efficiency.
Natural Indonesian Writing Skill
A specialized behavioral skill that enables AI agents to write Indonesian that sounds human, casual, and authentic rather than formal or robotic.
Featured*
