A hardened search integration that leverages the Perplexity API to provide AI-powered, grounded web answers with verifiable citations.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install perplexity-hardened
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install perplexity-hardened using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Perplexity Hardened is a robust search skill designed for AI agents within the Openclaw Skills ecosystem. It bridges the gap between static LLM knowledge and the live web by providing real-time, cited answers through the Perplexity API. Unlike standard search tools, this skill is armored with specific security guardrails to ensure data integrity and prevent the execution of malicious instructions found in search results.
By integrating this tool, developers can empower their agents to perform deep research, verify facts, and handle complex queries without sacrificing security. The skill focuses on returning high-fidelity information that remains local to the session, ensuring that Openclaw Skills users maintain complete control over their data flow.
To use this search capability within the Openclaw Skills framework, follow these steps:
export PERPLEXITY_API_KEY='your_api_key_here'
node scripts/search.mjs "current state of quantum computing"
The skill manages data primarily through standard output and optional JSON formatting. The following structure defines the response taxonomy:
| Attribute | Description |
|---|---|
| Query | The original search string(s) provided by the user. |
| Answer | The AI-generated response grounded in web search results. |
| Citations | A list of URLs or sources provided by the API, presented without modification. |
| JSON Output | Enabled via the --json flag to provide raw API metadata for programmatic use. |
All data is processed locally and is never piped to network-transmitting commands to prevent exfiltration.
--json flag to integrate search results directly into automated pipelines or other Openclaw Skills.Loading
A security-first AI code analysis skill that identifies bugs and visualizes logic flows through the LogicArt engine.

A lightweight watchdog and self-healing system designed to provide 24/7 health monitoring and automatic recovery for Openclaw environments.

A professional tool for systematically checking compatibility and managing risks when migrating embedded code between hardware platforms.

A powerful desktop automation skill that combines traditional Python libraries with advanced Vision AI to control any GUI environment.

A professional-grade reconnaissance tool for host discovery, service fingerprinting, and CVE matching without external dependencies.

An AI-optimized web search skill that leverages the Tavily API through AIsa's unified gateway to provide clean, relevant data for autonomous agents.








































