Enable real-time interaction through floating windows and modals for instant user feedback and data input without leaving the active application.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install realtime-interact-overlay
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 realtime-interact-overlay 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 Realtime Interact?
Realtime Interact is a specialized tool within the Openclaw Skills ecosystem designed to facilitate immediate human-in-the-loop interactions. Instead of forcing users to return to a chat interface, this skill injects native macOS dialogs or browser-based modals directly into the current workspace. It bridges the gap between autonomous execution and necessary human oversight for sensitive operations.
By leveraging this skill, developers can create more fluid AI workflows where confirmation, data entry, and selection happen in context. Whether managing local file systems or navigating complex web applications, Realtime Interact ensures that Openclaw Skills remain highly responsive and user-centric through modern, glassmorphism-styled UI components.
Realtime Interact Use Cases
- Confirming social media comments or posts before they are published to a live feed.
- Verifying destructive local actions such as file deletions or directory formatting.
- Securely collecting passwords or PINs during automated checkout or login sequences.
- Presenting dynamic selection lists for the user to choose from during multi-step tasks.
- Displaying custom HTML content for rich data visualization and immediate user feedback.
How Realtime Interact Works
- The AI agent identifies a trigger point in the workflow requiring human verification or input.
- The agent invokes the appropriate interaction method (confirm, input, select, or html) provided by the Openclaw Skills framework.
- The system detects the environment context—selecting a native macOS dialog for local tasks or a script-injected modal for browser tasks.
- A floating window appears on top of the user's active application with high-visibility styling.
- The user interacts with the window (clicking buttons, typing text, or selecting options).
- The skill captures the response and returns a structured JSON object to the AI agent to determine the next logical step.
Realtime Interact Setup
To utilize Realtime Interact effectively within your Openclaw Skills setup, follow these configuration steps:
- Permissions: For local macOS dialogs, you must grant Accessibility permissions. Navigate to System Settings > Security & Privacy > Accessibility and ensure your terminal/agent is authorized.
- Manual CLI Testing: You can test the dialog system using the included Python scripts:
python3 ~/.openclaw/skills/realtime-interact-1.0.0/scripts/macos_dialog.py \
--type confirm \
--title "Security Alert" \
--message "Confirm file deletion?"
- Browser Integration: For web-based modals, ensure the
browser_modal.pyscript has access to your active browser session via the agent's automation layer.
Realtime Interact Data Schema & Taxonomy
The skill communicates via a standardized JSON schema to ensure compatibility across different agent implementations.
| Attribute | Type | Description |
|---|---|---|
| result | string | The status of the interaction: 'confirm', 'cancel', 'timeout', or 'error'. |
| value | string/mixed | The actual data entered by the user or the selected option index. |
| message | string | Detailed error description in case of a 'error' result status. |
| timestamp | ISO8601 | The exact time the user submitted the interaction. |
Realtime Interact Advanced Features
- Multi-environment support: Switches automatically between system-level macOS dialogs and browser-level DOM modals.
- Glassmorphism UI: Modern, aesthetically pleasing browser overlays with blur effects and smooth transitions.
- Error & Timeout Handling: Built-in 60-second default timeout to prevent the AI agent from hanging indefinitely.
- Custom HTML Support: Capability to render complex forms or layouts directly within a floating browser window.
- Extensible Voice Hooks: Prepared architecture for future voice-based confirmation and TTS output within the Openclaw Skills ecosystem.
SKILL.md
Loading
Related Openclaw Skills
thesims
A social simulation world where AI agents with unique personalities interact, debate, trade, and build autonomous relationships.
the-soul-sims
An autonomous social simulation world where AI agents with unique SOUL.md personalities interact, trade, and build reputations.
The Sims
A shared social simulation world where AI agents with unique personalities interact, trade, and evolve autonomously.
The Agent Sims
A social simulation world where autonomous AI agents with unique personalities interact, trade, and build reputations in a shared digital environment.
Zhihu CLI
A comprehensive command-line tool for searching, reading, and managing interactions on the Zhihu platform.
FLUX Image Generation
A powerful command-line interface for generating professional-grade AI images using FLUX models via the inference.sh platform.
Featured*
