Agent skill

perplexity-search

A skill for performing web searches, research, and reasoning using the Perplexity API. Handles real-time web information retrieval, deep research analysis, and advanced reasoning tasks. Use when the user asks for web searches, research, or says things like "look up", "search for", "latest information", "research", or "analyze".

View SKILL.md on GitHub Repository
Stars 0
Forks 0

Install this agent skill to your Project

npx add-skill https://github.com/dotneet/claude-code-marketplace/tree/main/research/skills/perplexity-search

SKILL.md

Perplexity Search Skill

A skill for executing real-time web searches and research using the Perplexity API.

Purpose

This skill provides the following capabilities:

  1. perplexity_ask - Answer general questions (using sonar-pro model)
  2. perplexity_research - Deep research and comprehensive reports (using sonar-deep-research model)
  3. perplexity_reason - Advanced reasoning and analysis (using sonar-reasoning-pro model)
  4. perplexity_search - Retrieve web search results

When to Use

Use this skill in the following situations:

  • User needs up-to-date information
  • Received a question requiring web search
  • Asked to perform deep research or investigation
  • Complex analysis or reasoning is required
  • Keywords like "look up", "search for", "latest..." are included

Prerequisites

  • PERPLEXITY_API_KEY environment variable must be set
  • Internet connection must be available

Usage

Basic Usage

Use the scripts/perplexity_api.py script to call the API.

bash
# General questions (ask)
python3 scripts/perplexity_api.py ask "your question"

# Deep research (research)
python3 scripts/perplexity_api.py research "research topic"

# Advanced reasoning (reason)
python3 scripts/perplexity_api.py reason "reasoning task"

# Web search (search)
python3 scripts/perplexity_api.py search "search query" [--max-results 10] [--country JP]

Command Options

ask / research / reason

  • First argument: question, research topic, or reasoning task
  • --strip-thinking: Remove <think>...</think> tags to save context tokens (research/reason only)

search

  • First argument: search query
  • --max-results: Maximum number of results to return (1-20, default: 10)
  • --max-tokens-per-page: Maximum tokens per page (256-2048, default: 1024)
  • --country: ISO code for regional results (e.g., JP, US, GB)

Workflow

Standard Search Flow

  1. Analyze the user's question and select the appropriate tool

    • Simple questions → ask
    • Deep research → research
    • Complex analysis → reason
    • Information gathering → search

  2. Execute the script to call the API

  3. Present results to the user, citing sources when available

Tool Selection Guidelines

Use Case Tool Description
Current weather, news ask When quick answers are needed
Technical topic research research When comprehensive analysis is needed
Complex problem analysis reason When logical reasoning is needed
Collecting sources search When URLs or snippets are needed

API Details

For detailed API specifications, see references/api_reference.md.

Troubleshooting

  • API Key Error: Verify the PERPLEXITY_API_KEY environment variable
  • Timeout: Increase PERPLEXITY_TIMEOUT_MS (default: 300000ms)
  • Proxy Issues: Set PERPLEXITY_PROXY or HTTPS_PROXY

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

dotneet/claude-code-marketplace

prompt-improver

A skill for improving prompts by applying general LLM/agent best practices. When the user provides a prompt, this skill outputs an improved version, identifies missing information, and provides specific improvement points. Use when the user asks to "improve this prompt", "review this prompt", or "make this prompt better".

0 0
Explore
dotneet/claude-code-marketplace

suggest-agent-rules

Suggest agent rules analyzing the session history and the current repository.

0 0
Explore
dotneet/claude-code-marketplace

context7

A skill for retrieving the latest library documentation using Context7. Use when the user asks about how to use a library, requests code examples, or instructs to "use context7". Prevents hallucinations based on outdated training data and provides up-to-date API information.

0 0
Explore
dotneet/claude-code-marketplace

code-modularization-evaluator

Evaluate and improve code modularization using the Balanced Coupling Model. Analyzes coupling strength, connascence types, and distance to identify refactoring opportunities and architectural improvements. Use when reviewing code architecture, refactoring modules, or designing new systems.

0 0
Explore
dotneet/claude-code-marketplace

typescript-react-reviewer

Expert code reviewer for TypeScript + React 19 applications. Use when reviewing React code, identifying anti-patterns, evaluating state management, or assessing code maintainability. Triggers: code review requests, PR reviews, React architecture evaluation, identifying code smells, TypeScript type safety checks, useEffect abuse detection, state management review.

0 0
Explore
dotneet/claude-code-marketplace

google-calendar

Accesses the Google Calendar and Tasks API. Use this skill to view or manage calendars, events, schedules, or Google tasks.

0 0
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results