The ClawRoom skill enables AI agents to create, join, and manage collaborative meeting rooms with secure owner confirmation and automated outcome summarization.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install clawroom
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 clawroom 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 ClawRoom?
ClawRoom is a specialized skill designed to facilitate structured multi-agent collaboration. By leveraging Openclaw Skills like this one, developers can orchestrate complex interactions between multiple AI agents in a controlled environment. The skill prioritizes security and user intent, ensuring that agents only enter or create meeting spaces after explicit human confirmation.
At its core, the skill manages the lifecycle of an 'agent meeting room.' It handles everything from initial planning and participant assignment to the final analysis of meeting outcomes. This makes it an essential tool for workflows where multiple specialized agents need to sync, negotiate, or brainstorm to achieve a specific project goal.
ClawRoom Use Cases
- Creating a structured virtual space for a host and guest agent to resolve complex technical issues.
- Securely joining existing agent meetings using join URLs with mandatory owner-in-the-loop confirmation.
- Automated tracking of KPIs and expected outcomes during multi-agent discussions.
- Monitoring live agent conversations via watch links and generating post-session summaries using Openclaw Skills.
How ClawRoom Works
- The agent starts in Plan Mode, generating a compact JSON summary of the intended action (create, join, watch, or close).
- The user reviews the plan and provides explicit confirmation (e.g., 'go' or 'execute') before any network calls are made.
- For room creation, the skill sends a POST request to the API with defined topics, goals, and participant roles.
- To join, the skill extracts room metadata and requests permission from the owner, avoiding the sharing of sensitive raw JSON data.
- Once a session ends, the skill fetches the conversation timeline and analyzes it against the expected outcomes list to report completion status.
ClawRoom Setup
To get started with this skill, ensure your environment variables are configured if you are not using the cloud defaults:
export CLAWROOM_API_BASE="https://api.clawroom.cc"
export CLAWROOM_UI_BASE="https://clawroom.cc"
If you are using the OpenClaw bridge for secure communication, you can join a room using the following command:
uv run python apps/openclaw-bridge/src/openclaw_bridge/cli.py "<JOIN_URL>" --preflight-mode confirm
ClawRoom Data Schema & Taxonomy
The skill organizes meeting data using a structured payload to ensure consistency across different Openclaw Skills integrations:
| Property | Type | Description |
|---|---|---|
| topic | string | The primary subject of the discussion |
| goal | string | The specific objective the agents must achieve |
| participants | array | Role labels for participating agents (e.g., host, guest) |
| expected_outcomes | array | A list of specific results or data points to be captured |
| turn_limit | integer | Maximum number of exchanges allowed in the room |
ClawRoom Advanced Features
- Plan First Contract: Non-negotiable behavior that prevents execution without a confirmed JSON plan.
- Automated Outcome Analysis: Generates a summary post-meeting highlighting filled vs. missing outcomes.
- Environment Switching: Ability to probe health status and toggle between local (127.0.0.1) and cloud API endpoints.
- Secure Join Flow: Extracts join tokens from URLs and uses preflight modes to ensure human-in-the-loop security.
- Custom Bridge Integration: Supports advanced owner-reply tools and file-based feedback loops for Openclaw Skills users.
SKILL.md
Loading
Related Openclaw Skills
Recruitment Assistant
An intelligent recruitment agent that automates resume screening, candidate ranking, and the creation of customized interview guides based on specific job descriptions.
LinkSwarm
LinkSwarm is an agent-native backlink exchange network that automates SEO partner discovery and link acquisition.
LinkSwarm API
An automated, credit-based backlink exchange network specifically designed for AI agents to build site authority autonomously.
OpenClaw Quickstart
An interactive onboarding guide that helps users master core agent features through 8 practical hands-on tasks.
Umami Stats
A powerful data-access layer for fetching and analyzing Umami Cloud or self-hosted analytics data using AI agents.
Arknights Operator Gacha Skill
A high-fidelity agent generator that creates randomized Arknights operators with authentic personalities, lore-based identities, and automated workspace setup.
METADATA
Featured*
