crewai-aisearchapi 1.0.7

AI Search API integration for CrewAI - Enable intelligent web search in your crews

AI Search API for CrewAI - Developer & SEO Guide

A Python integration for CrewAI that connects your agents to the AI Search API.
It enables semantic search, contextual queries, SEO‑friendly content generation, and intelligent answers with sources.

👉 To start, get your free API key from the AI Search API dashboard.

---

Features

• 🔍 AI-Powered Web Search: Let CrewAI agents search the web with natural language
• 🎯 Context Awareness: Add history messages for better, more relevant answers
• ⚡ Seamless CrewAI Integration: Works out of the box with Agent, Task, and Crew
• 🛡️ Strong Typing: Includes dataclasses and type hints for safe development
• 💡 SEO Use Cases: Generate SEO‑optimized content, product descriptions, FAQs
• 📊 Sources: Always returns sources for transparency
• 🖥️ Local LLM Support: Combine with free Ollama models for offline reasoning

---

Installation

pip install crewai-aisearchapi

---

Quick Start (with Ollama + CrewAI)

from crewai import Agent, Task, Crew, Process, LLM
from crewai_aisearchapi import AISearchTool

# Use a local free model (Ollama)
llm = LLM(
    model="ollama/llama3.2:3b",            # or llama3.1:8b
    base_url="http://localhost:11434",     # Ollama default
    temperature=0.2,
)

# Add the AI Search tool
tool = AISearchTool(api_key="your-api-key")

agent = Agent(
    role="Researcher",
    goal="Short, correct answers with sources when available.",
    backstory="Careful and concise.",
    tools=[tool],
    llm=llm,
    verbose=True,
)

task = Task(
    description="Answer: '{question}'. Keep it short.",
    expected_output="2–4 sentences.",
    agent=agent,
    markdown=True,
)

crew = Crew(agents=[agent], tasks=[task], process=Process.sequential, verbose=True)

if __name__ == "__main__":
    print(crew.kickoff(inputs={"question": "What is RLHF in AI?"}))

---

Configuration Options

from crewai_aisearchapi import AISearchTool, AISearchToolConfig

config = AISearchToolConfig(
    default_response_type="markdown",  # or "text"
    include_sources=True,
    timeout=30,
    verbose=True
)

tool = AISearchTool(api_key="your-key", config=config)

---

Contextual Search

You can pass multiple context messages (all must be role "user"):

result = tool.run({
    "prompt": "Summarize quantum dots for displays.",
    "context": [
        {"role": "user", "content": "Be concise."},
        {"role": "user", "content": "Focus on consumer TVs."}
    ],
    "response_type": "markdown"
})

---

SEO Use Cases

The AI Search API is not just for research — it’s also great for SEO content generation:

• 📈 Write product descriptions with live data & sources
• 📝 Generate blog posts enriched with context and citations
• ❓ Build FAQ sections that adapt to user intent
• 🌍 Research competitor content and summarize strengths/weaknesses
• ⚙️ Automate content briefs for writers using CrewAI agents

Example:

seo_agent = Agent(
    role="SEO Writer",
    goal="Generate SEO-optimized product descriptions.",
    backstory="Expert in e-commerce SEO content.",
    tools=[tool],
    llm=llm
)

seo_task = Task(
    description="Write a 150-word SEO product description for '{product}'",
    expected_output="SEO-rich product description with 2 citations.",
    agent=seo_agent
)

---

Handling Responses

The tool automatically formats results:

The current Bitcoin price is approximately $45,000 USD...

**Sources:**
- [1] https://coinmarketcap.com
- [2] https://finance.yahoo.com

*Response time: 150ms*

---

Environment Variables

export AISEARCH_API_KEY="your-api-key"

In Python:

import os
from crewai_aisearchapi import AISearchTool

tool = AISearchTool(api_key=os.getenv("AISEARCH_API_KEY"))

---

Development (for contributors)

git clone https://github.com/aisearchapi/crewai-aisearchapi.git
cd crewai-aisearchapi
python -m venv venv
source venv/bin/activate
pip install -e ".[dev,test]"
pytest

---

License

MIT License - see the LICENSE file.

---

Support

• Dashboard & API Key: AI Search API Dashboard
• Docs: docs.aisearchapi.io
• Homepage: aisearchapi.io
• Issues: GitHub Issues