Cloak enables secure, one-time secret sharing between humans and AI agents via encrypted, self-destructing links.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install cloak
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 cloak 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 Cloak?
Cloak is a specialized security utility designed to facilitate the safe exchange of sensitive information like API keys, tokens, and passwords. As a vital component for Openclaw Skills, it ensures that secrets are never exposed in plain text within chat logs or LLM conversation histories. By generating unique, single-use URLs, Cloak allows users and agents to transmit credentials that automatically self-destruct after a single retrieval or a specified expiration period.
Built for high-security environments, Cloak prioritizes privacy by using header-based key transmission rather than query parameters. This ensures that sensitive data remains out of server logs and browser histories, providing a robust layer of protection for developers and automated workflows that require temporary access to private resources.
Cloak Use Cases
- Sharing temporary API tokens with an agent to perform cloud infrastructure tasks.
- Transmitting database credentials for local environment setup without manual entry.
- Securely passing environment variables into a CI/CD pipeline or configuration file.
- Sending one-time passwords or recovery codes to a remote team member or agent.
How Cloak Works
- The sender creates a secret by sending a POST request to the Cloak API containing the payload.
- The API returns a unique ID, a decryption key, and a self-destructing URL containing both.
- The recipient accesses the secret by sending a GET request with the decryption key provided in the X-Cloak-Key header.
- Once the secret is retrieved, the Cloak server immediately and permanently deletes the data.
- The agent follows strict security rules to pipe the retrieved secret directly to its destination, such as an environment variable or a config file, without ever displaying it in the chat interface.
Cloak Setup
To integrate Cloak into your environment for use with Openclaw Skills, ensure you have curl and jq installed. No complex installation is required as it interacts via a standard REST API.
Example of creating a secret via CLI:
curl -s -X POST https://cloak.opsy.sh/api/secrets \
-H "Content-Type: application/json" \
-d '{"secret":"YOUR_SECRET_HERE"}'
Example of retrieving and piping a secret to an environment variable:
export API_KEY=$(curl -s -H "X-Cloak-Key: YOUR_KEY" "https://cloak.opsy.sh/api/secrets/YOUR_ID" | jq -r .secret)
Cloak Data Schema & Taxonomy
Cloak utilizes a ephemeral data model where secrets are associated with unique identifiers and specific TTL (Time To Live) settings. The Openclaw Skills integration expects the following data structure:
| Field | Type | Description |
|---|---|---|
| id | String | Unique path identifier for the secret retrieval. |
| key | String | The decryption key required in the X-Cloak-Key header. |
| url | String | The full shareable link (Format: /s/{id}#{key}). |
| expiresAt | Integer | Unix timestamp when the secret will be automatically purged. |
| expiresIn | Integer | Optional custom TTL in seconds (default: 86400). |
Cloak Advanced Features
- Custom TTL support allowing secrets to persist from a few seconds up to 7 days.
- Programmatic deletion capability to revoke a secret before it is ever viewed.
- Native support for piping secrets directly into shell environments or .env files to bypass terminal history.
- Strict no-logging architecture that prevents sensitive keys from appearing in URL parameters.
SKILL.md
Loading
Related Openclaw Skills
Gemini CLI
A powerful command-line interface for Google Gemini models, designed for codebase analysis, AI-driven code generation, and automated developer workflows.
AI Data Visualization Tool
An intelligent automation skill that transforms complex datasets into high-impact visual reports and actionable insights using advanced AI processing.
AI Content Rewriting Tool
A high-performance AI agent skill designed to automate content rewriting and multi-language transformation for professional teams.
nano-pdf
A command-line interface for applying precise PDF edits through natural language instructions.
Repo Surveyor
A specialized skill for manual repository orientation and task context generation when automated mapping tools are unavailable.
Lingua Universale
A mathematical verification engine for AI agent-to-agent communication based on session type protocols.
Featured*
