Brandon/cre 130 pipeline project structure (#1066)

* WIP. Procedure appears to be working well. Working on mocking properly for tests

* All tests are passing now

* rshift working

* Add back in Gui's tool_usage fix

* WIP

* Going to start refactoring for pipeline_output

* Update terminology

* new pipeline flow with traces and usage metrics working. need to add more tests and make sure PipelineOutput behaves likew CrewOutput

* Fix pipelineoutput to look more like crewoutput and taskoutput

* Implemented additional tests for pipeline. One test is failing. Need team support

* Update docs for pipeline

* Update pipeline to properly process input and ouput dictionary

* Update Pipeline docs

* Add back in commentary at top of pipeline file

* Starting to work on router

* Drop router for now. will add in separately

* In the middle of fixing router. A ton of circular dependencies. Moving over to a new design.

* WIP.

* Fix circular dependencies and updated PipelineRouter

* Add in Eduardo feedback. Still need to add in more commentary describing the design decisions for pipeline

* Add developer notes to explain what is going on in pipelines.

* Add doc strings

* Fix missing rag datatype

* WIP. Converting usage metrics from a dict to an object

* Fix tests that were checking usage metrics

* Drop todo

* Fix 1 type error in pipeline

* Update pipeline to use UsageMetric

* Add missing doc string

* WIP.

* Change names

* Rename variables based on joaos feedback

* Fix critical circular dependency issues. Now needing to fix trace issue.

* Tests working now!

* Add more tests which showed underlying issue with traces

* Fix tests

* Remove overly complicated test

* Add router example to docs

* Clean up end of docs

* Clean up docs

* Working on creating Crew templates and pipeline templates

* WIP.

* WIP

* Fix poetry install from templates

* WIP

* Restructure

* changes for lorenze

* more todos

* WIP: create pipelines cli working

* wrapped up router

* ignore mypy src on templates

* ignored signature of copy

* fix all verbose

* rm print statements

* brought back correct folders

* fixes missing folders and then rm print statements

* fixed tests

* fixed broken test

* fixed type checker

* fixed type ignore

* ignore types for templates

* needed

* revert

* exclude only required

* rm type errors on templates

* rm excluding type checks for template files on github action

* fixed missing quotes

---------

Co-authored-by: Brandon Hancock <brandon@brandonhancock.io>
This commit is contained in:
Lorenze Jay
2024-08-09 14:13:29 -07:00
committed by GitHub
parent 6583f31459
commit 62f5b2fb2e
87 changed files with 5435 additions and 571 deletions

View File

@@ -0,0 +1,2 @@
.env
__pycache__/

View File

@@ -0,0 +1,57 @@
# {{crew_name}} Crew
Welcome to the {{crew_name}} Crew project, powered by [crewAI](https://crewai.com). This template is designed to help you set up a multi-agent AI system with ease, leveraging the powerful and flexible framework provided by crewAI. Our goal is to enable your agents to collaborate effectively on complex tasks, maximizing their collective intelligence and capabilities.
## Installation
Ensure you have Python >=3.10 <=3.13 installed on your system. This project uses [Poetry](https://python-poetry.org/) for dependency management and package handling, offering a seamless setup and execution experience.
First, if you haven't already, install Poetry:
```bash
pip install poetry
```
Next, navigate to your project directory and install the dependencies:
1. First lock the dependencies and then install them:
```bash
poetry lock
```
```bash
poetry install
```
### Customizing
**Add your `OPENAI_API_KEY` into the `.env` file**
- Modify `src/{{folder_name}}/config/agents.yaml` to define your agents
- Modify `src/{{folder_name}}/config/tasks.yaml` to define your tasks
- Modify `src/{{folder_name}}/crew.py` to add your own logic, tools and specific args
- Modify `src/{{folder_name}}/main.py` to add custom inputs for your agents and tasks
## Running the Project
To kickstart your crew of AI agents and begin task execution, run this from the root folder of your project:
```bash
poetry run {{folder_name}}
```
This command initializes the {{name}} Crew, assembling the agents and assigning them tasks as defined in your configuration.
This example, unmodified, will run the create a `report.md` file with the output of a research on LLMs in the root folder.
## Understanding Your Crew
The {{name}} Crew is composed of multiple AI agents, each with unique roles, goals, and tools. These agents collaborate on a series of tasks, defined in `config/tasks.yaml`, leveraging their collective skills to achieve complex objectives. The `config/agents.yaml` file outlines the capabilities and configurations of each agent in your crew.
## Support
For support, questions, or feedback regarding the {{crew_name}} Crew or crewAI.
- Visit our [documentation](https://docs.crewai.com)
- Reach out to us through our [GitHub repository](https://github.com/joaomdmoura/crewai)
- [Join our Discord](https://discord.com/invite/X4JWnZnxPb)
- [Chat with our docs](https://chatg.pt/DWjSBZn)
Let's create wonders together with the power and simplicity of crewAI.

View File

@@ -0,0 +1,19 @@
researcher:
role: >
{topic} Senior Data Researcher
goal: >
Uncover cutting-edge developments in {topic}
backstory: >
You're a seasoned researcher with a knack for uncovering the latest
developments in {topic}. Known for your ability to find the most relevant
information and present it in a clear and concise manner.
reporting_analyst:
role: >
{topic} Reporting Analyst
goal: >
Create detailed reports based on {topic} data analysis and research findings
backstory: >
You're a meticulous analyst with a keen eye for detail. You're known for
your ability to turn complex data into clear and concise reports, making
it easy for others to understand and act on the information you provide.

View File

@@ -0,0 +1,16 @@
research_task:
description: >
Conduct a thorough research about {topic}
Make sure you find any interesting and relevant information given
the current year is 2024.
expected_output: >
A list with 10 bullet points of the most relevant information about {topic}
agent: researcher
reporting_task:
description: >
Review the context you got and expand each topic into a full section for a report.
Make sure the report is detailed and contains any and all relevant information.
expected_output: >
A fully fledge reports with a title, mains topics, each with a full section of information.
agent: reporting_analyst

View File

@@ -0,0 +1,58 @@
from pydantic import BaseModel
from crewai import Agent, Crew, Process, Task
from crewai.project import CrewBase, agent, crew, task
# Uncomment the following line to use an example of a custom tool
# from demo_pipeline.tools.custom_tool import MyCustomTool
# Check our tools documentations for more information on how to use them
# from crewai_tools import SerperDevTool
class ResearchReport(BaseModel):
"""Research Report"""
title: str
body: str
@CrewBase
class ResearchCrew():
"""Research Crew"""
agents_config = 'config/agents.yaml'
tasks_config = 'config/tasks.yaml'
@agent
def researcher(self) -> Agent:
return Agent(
config=self.agents_config['researcher'],
verbose=True
)
@agent
def reporting_analyst(self) -> Agent:
return Agent(
config=self.agents_config['reporting_analyst'],
verbose=True
)
@task
def research_task(self) -> Task:
return Task(
config=self.tasks_config['research_task'],
)
@task
def reporting_task(self) -> Task:
return Task(
config=self.tasks_config['reporting_task'],
output_pydantic=ResearchReport
)
@crew
def crew(self) -> Crew:
"""Creates the Research Crew"""
return Crew(
agents=self.agents, # Automatically created by the @agent decorator
tasks=self.tasks, # Automatically created by the @task decorator
process=Process.sequential,
verbose=True,
)

View File

@@ -0,0 +1,51 @@
from crewai import Agent, Crew, Process, Task
from crewai.project import CrewBase, agent, crew, task
# Uncomment the following line to use an example of a custom tool
# from {{folder_name}}.tools.custom_tool import MyCustomTool
# Check our tools documentations for more information on how to use them
# from crewai_tools import SerperDevTool
@CrewBase
class WriteLinkedInCrew():
"""Research Crew"""
agents_config = 'config/agents.yaml'
tasks_config = 'config/tasks.yaml'
@agent
def researcher(self) -> Agent:
return Agent(
config=self.agents_config['researcher'],
verbose=True
)
@agent
def reporting_analyst(self) -> Agent:
return Agent(
config=self.agents_config['reporting_analyst'],
verbose=True
)
@task
def research_task(self) -> Task:
return Task(
config=self.tasks_config['research_task'],
)
@task
def reporting_task(self) -> Task:
return Task(
config=self.tasks_config['reporting_task'],
output_file='report.md'
)
@crew
def crew(self) -> Crew:
"""Creates the {{crew_name}} crew"""
return Crew(
agents=self.agents, # Automatically created by the @agent decorator
tasks=self.tasks, # Automatically created by the @task decorator
process=Process.sequential,
verbose=True,
)

View File

@@ -0,0 +1,14 @@
x_writer_agent:
role: >
Expert Social Media Content Creator specializing in short form written content
goal: >
Create viral-worthy, engaging short form posts that distill complex {topic} information
into compelling 280-character messages
backstory: >
You're a social media virtuoso with a particular talent for short form content. Your posts
consistently go viral due to your ability to craft hooks that stop users mid-scroll.
You've studied the techniques of social media masters like Justin Welsh, Dickie Bush,
Nicolas Cole, and Shaan Puri, incorporating their best practices into your own unique style.
Your superpower is taking intricate {topic} concepts and transforming them into
bite-sized, shareable content that resonates with a wide audience. You know exactly
how to structure a post for maximum impact and engagement.

View File

@@ -0,0 +1,22 @@
write_x_task:
description: >
Using the research report provided, create an engaging short form post about {topic}.
Your post should have a great hook, summarize key points, and be structured for easy
consumption on a digital platform. The post must be under 280 characters.
Follow these guidelines:
1. Start with an attention-grabbing hook
2. Condense the main insights from the research
3. Use clear, concise language
4. Include a call-to-action or thought-provoking question if space allows
5. Ensure the post flows well and is easy to read quickly
Here is the title of the research report you will be using
Title: {title}
Research:
{body}
expected_output: >
A compelling X post under 280 characters that effectively summarizes the key findings
about {topic}, starts with a strong hook, and is optimized for engagement on the platform.
agent: x_writer_agent

View File

@@ -0,0 +1,36 @@
from crewai import Agent, Crew, Process, Task
from crewai.project import CrewBase, agent, crew, task
# Uncomment the following line to use an example of a custom tool
# from demo_pipeline.tools.custom_tool import MyCustomTool
# Check our tools documentations for more information on how to use them
# from crewai_tools import SerperDevTool
@CrewBase
class WriteXCrew:
"""Research Crew"""
agents_config = "config/agents.yaml"
tasks_config = "config/tasks.yaml"
@agent
def x_writer_agent(self) -> Agent:
return Agent(config=self.agents_config["x_writer_agent"], verbose=True)
@task
def write_x_task(self) -> Task:
return Task(
config=self.tasks_config["write_x_task"],
)
@crew
def crew(self) -> Crew:
"""Creates the Write X Crew"""
return Crew(
agents=self.agents, # Automatically created by the @agent decorator
tasks=self.tasks, # Automatically created by the @task decorator
process=Process.sequential,
verbose=True,
)

View File

@@ -0,0 +1,26 @@
#!/usr/bin/env python
import asyncio
from {{folder_name}}.pipelines.pipeline import {{pipeline_name}}Pipeline
async def run():
"""
Run the pipeline.
"""
inputs = [
{"topic": "AI wearables"},
]
pipeline = {{pipeline_name}}Pipeline()
results = await pipeline.kickoff(inputs)
# Process and print results
for result in results:
print(f"Raw output: {result.raw}")
if result.json_dict:
print(f"JSON output: {result.json_dict}")
print("\n")
def main():
asyncio.run(run())
if __name__ == "__main__":
main()

View File

@@ -0,0 +1,87 @@
"""
This pipeline file includes two different examples to demonstrate the flexibility of crewAI pipelines.
Example 1: Two-Stage Pipeline
-----------------------------
This pipeline consists of two crews:
1. ResearchCrew: Performs research on a given topic.
2. WriteXCrew: Generates an X (Twitter) post based on the research findings.
Key features:
- The ResearchCrew's final task uses output_json to store all research findings in a JSON object.
- This JSON object is then passed to the WriteXCrew, where tasks can access the research findings.
Example 2: Two-Stage Pipeline with Parallel Execution
-------------------------------------------------------
This pipeline consists of three crews:
1. ResearchCrew: Performs research on a given topic.
2. WriteXCrew and WriteLinkedInCrew: Run in parallel, using the research findings to generate posts for X and LinkedIn, respectively.
Key features:
- Demonstrates the ability to run multiple crews in parallel.
- Shows how to structure a pipeline with both sequential and parallel stages.
Usage:
- To switch between examples, comment/uncomment the respective code blocks below.
- Ensure that you have implemented all necessary crew classes (ResearchCrew, WriteXCrew, WriteLinkedInCrew) before running.
"""
# Common imports for both examples
from crewai import Pipeline
# Uncomment the crews you need for your chosen example
from ..crews.research_crew.research_crew import ResearchCrew
from ..crews.write_x_crew.write_x_crew import WriteXCrew
# from .crews.write_linkedin_crew.write_linkedin_crew import WriteLinkedInCrew # Uncomment for Example 2
# EXAMPLE 1: Two-Stage Pipeline
# -----------------------------
# Uncomment the following code block to use Example 1
class {{pipeline_name}}Pipeline:
def __init__(self):
# Initialize crews
self.research_crew = ResearchCrew().crew()
self.write_x_crew = WriteXCrew().crew()
def create_pipeline(self):
return Pipeline(
stages=[
self.research_crew,
self.write_x_crew
]
)
async def kickoff(self, inputs):
pipeline = self.create_pipeline()
results = await pipeline.kickoff(inputs)
return results
# EXAMPLE 2: Two-Stage Pipeline with Parallel Execution
# -------------------------------------------------------
# Uncomment the following code block to use Example 2
# @PipelineBase
# class {{pipeline_name}}Pipeline:
# def __init__(self):
# # Initialize crews
# self.research_crew = ResearchCrew().crew()
# self.write_x_crew = WriteXCrew().crew()
# self.write_linkedin_crew = WriteLinkedInCrew().crew()
# @pipeline
# def create_pipeline(self):
# return Pipeline(
# stages=[
# self.research_crew,
# [self.write_x_crew, self.write_linkedin_crew] # Parallel execution
# ]
# )
# async def run(self, inputs):
# pipeline = self.create_pipeline()
# results = await pipeline.kickoff(inputs)
# return results

View File

@@ -0,0 +1,17 @@
[tool.poetry]
name = "{{folder_name}}"
version = "0.1.0"
description = "{{name}} using crewAI"
authors = ["Your Name <you@example.com>"]
[tool.poetry.dependencies]
python = ">=3.10,<=3.13"
crewai = { extras = ["tools"], version = "^0.46.0" }
asyncio = "*"
[tool.poetry.scripts]
{{folder_name}} = "{{folder_name}}.main:main"
[build-system]
requires = ["poetry-core"]
build-backend = "poetry.core.masonry.api"

View File

@@ -0,0 +1,12 @@
from crewai_tools import BaseTool
class MyCustomTool(BaseTool):
name: str = "Name of my tool"
description: str = (
"Clear description for what this tool is useful for, you agent will need this information to use it."
)
def _run(self, argument: str) -> str:
# Implementation goes here
return "this is an example of a tool output, ignore it and move along."