Web Search (Perplexity) for Openclaw

A high-performance search and content retrieval skill for AI agents powered by the Perplexity API and local markdown extraction.

aligurelli
v1.5.0
Feb 17, 2026
6
2.3k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install perplexity-web-search

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install perplexity-web-search 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 Web Search (Perplexity)?

The Web Search skill integrates Perplexity's powerful search models into your AI agent workflow, allowing for real-time research and synthesized findings with accurate citations. It serves as a robust bridge between local coding environments and the live internet, enabling the retrieval of up-to-the-minute information, technical documentation, and news.

By leveraging this component within Openclaw Skills, developers can equip their agents with the ability to answer complex queries and extract clean, readable markdown from specific URLs. This eliminates the need for manual browsing and ensures that your AI has the most relevant context available for any given task.

Web Search (Perplexity) Use Cases

  • Researching the latest software updates and technical documentation in real-time.
  • Monitoring social sentiment by searching platforms like Reddit, Twitter, and Quora for user feedback.
  • Fetching and summarizing the contents of specific web pages for local analysis without manual copy-pasting.
  • Retrieving location-specific data such as weather, news, or market trends using country-specific filters.

How Web Search (Perplexity) Works

  1. The agent receives a search or fetch request through the Openclaw Skills interface.
  2. For search tasks, the skill calls the Perplexity API using the configured model, such as sonar or sonar-pro.
  3. Perplexity processes the query and returns a synthesized answer accompanied by web citations.
  4. For fetch tasks, the skill retrieves the HTML content of a specified URL and extracts it into markdown or text format locally.
  5. The refined data is returned to the AI agent to be used for further reasoning or coding tasks.

Web Search (Perplexity) Setup

To get started with this skill in Openclaw Skills, ensure you have a valid Perplexity API key. Add your key to your environment variables:

export PERPLEXITY_API_KEY="your_api_key_here"

Configure the tool in your configuration file to define the model and provider:

{
  "tools": {
    "web": {
      "search": {
        "provider": "perplexity",
        "perplexity": {
          "apiKey": "<your-perplexity-api-key>",
          "model": "sonar"
        }
      }
    }
  }
}

Web Search (Perplexity) Data Schema & Taxonomy

The skill manages search parameters and content extraction through a structured schema designed for Openclaw Skills efficiency:

Parameter Description Configuration
query The search string to process Required String
count Number of results to return 1 to 10
freshness Time-based result filter pd, pw, pm, py
extractMode Format of fetched content markdown or text
country Regional search filtering ISO 2-letter codes

Web Search (Perplexity) Advanced Features

  • Support for multi-step reasoning models like sonar-pro and sonar-reasoning-pro for deep-dive research.
  • Native integration of social platform results from Reddit and Quora via natural language queries.
  • Targeted discovery using the site: operator for platform-specific information retrieval.
  • Efficient local content extraction via the web_fetch tool to minimize token usage and API latency.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*