Quick Start: No LLM Required

Fastest Way to Test AgentLantern

Follow this guide to create and test a minimal CrewAI project without needing any LLM or API keys.

Step 1: Install CrewAI

pip install crewai

Step 2: Create a Simple Project

crewai create my_test_project
cd my_test_project

Step 3: Modify agents.yaml (No API Keys Needed)

Edit src/my_test_project/config/agents.yaml:

researcher:
  role: Data Researcher
  goal: Research and analyze topics
  backstory: You are an expert at finding and analyzing information.
  tools: []
  verbose: true

writer:
  role: Content Writer
  goal: Write high quality content
  backstory: You are an excellent writer who creates engaging content.
  tools: []
  verbose: true

analyst:
  role: Data Analyst
  goal: Analyze and interpret data
  backstory: You are skilled at analyzing patterns and insights.
  tools: []
  verbose: true

Step 4: Modify tasks.yaml

Edit src/my_test_project/config/tasks.yaml:

research_task:
  description: Research the topic thoroughly
  expected_output: Comprehensive research document
  agent: researcher

writing_task:
  description: Write an article based on research
  expected_output: Well-written article
  agent: writer
  
analysis_task:
  description: Analyze the written content
  expected_output: Analysis report
  agent: analyst

Step 5: Modify crew.py

Edit src/my_test_project/crew.py - remove LLM initialization:

from crewai import Agent, Task, Crew

class MyTestProjectCrew:
    """Your test crew implementation without LLM"""

    def __init__(self):
        """Initialize the crew without LLM dependencies"""
        pass

    def setup_agents(self):
        """Setup agents (no LLM calls)"""
        self.researcher = Agent(
            role="Data Researcher",
            goal="Research topics",
            backstory="Expert researcher",
            verbose=True,
            allow_delegation=False
        )

        self.writer = Agent(
            role="Content Writer",
            goal="Write content",
            backstory="Expert writer",
            verbose=True,
            allow_delegation=False
        )

        self.analyst = Agent(
            role="Data Analyst",
            goal="Analyze data",
            backstory="Expert analyst",
            verbose=True,
            allow_delegation=False
        )

    def setup_tasks(self):
        """Setup tasks"""
        self.research_task = Task(
            description="Research the topic",
            expected_output="Research document",
            agent=self.researcher
        )

        self.writing_task = Task(
            description="Write article",
            expected_output="Article",
            agent=self.writer
        )

        self.analysis_task = Task(
            description="Analyze content",
            expected_output="Analysis",
            agent=self.analyst
        )

    def crew(self) -> Crew:
        """Create the crew"""
        self.setup_agents()
        self.setup_tasks()

        return Crew(
            agents=[self.researcher, self.writer, self.analyst],
            tasks=[self.research_task, self.writing_task, self.analysis_task],
            verbose=True
        )

Step 6: Install AgentLantern

pip install "agentlantern @ git+https://github.com/brellsanwouo/agentlantern.git"

Step 7: Generate Documentation

# From your project root
lantern docs

You should see:

Generated AgentLantern docs for my_test_project
- /path/to/project/docs/overview.md
- /path/to/project/docs/architecture.md
- /path/to/project/docs/agents.md
- /path/to/project/docs/tasks.md
- ... and more

Step 8: View Documentation

lantern web

Open your browser to http://localhost:9000 and explore:

• Agents - See your three agents (Researcher, Writer, Analyst)
• Tasks - See your three tasks with dependencies
• Architecture - Visual overview of your crew
• Runbook - How to execute your crew

What You Just Did

✅ Created a minimal CrewAI project
✅ Defined agents and tasks (no LLM needed)
✅ Generated documentation with AgentLantern
✅ Viewed beautiful documentation locally

That's It!

No API keys, no LLM setup, no complex configuration. You now have working documentation for a CrewAI project.

Next Steps

• Learn Usage Patterns
• Explore Examples
• Check API Reference
• Add more agents and tasks to your project
• Explore the generated documentation files
• Try deploying to GitHub Pages

Tips

• You can have as many agents and tasks as you want
• No LLM = no costs
• Perfect for testing and learning
• Documentation generates instantly
• Export docs anywhere - no dependencies needed