Slack Research Agent

Build AI Research Agent using Langchain and integrate it with Slack

About this project

This project demonstrates how to create a powerful tool designed to provide users with quick and accurate summaries of information about the company website mentioned in Slack channels. Leveraging LangChain, this agent uses Wikipedia and web search tools(I have used Tavily because it's free and easy to generate API keys, you could replace it with Bing, Google, etc.) to gather and summarize relevant data efficiently.

Prerequisites

Python 3.9 or higher installed on your system.
Slack account.

Slack Setup

It is important to configure your Slack bot correctly for this code to work.

Step 1: Create a Slack Account

Sign Up or Log In:
- Visit Slack app page.
- If you don't have an account, sign up. Otherwise, log in with your existing credentials.

Step 2: Configure Slack

Create a Slack App:
- Click Create New App
- Select From Scratch
- Enter an App Name and select the workspace(You can choose Slack if you do not have a workspace)
- Click Create App Once the app is created successfully you will be redirected to Basic Information page.
Configure Scope:
- Navigate to OAuth & Permissions under Features section.
- Go to Bot Token Scopes in the Scope section, click Add an OAuth Scope
- Add scopes "chat:write" and "chat:write.public"
Install App:
- Click Install to Workspace button on the OAuth & Permissions page under OAuth Tokens for Your Workspace section.
- Click Allow and you will be redirected to the OAuth & Permissions page again
- Copy the Bot User OAuth Token and save it in your .env file as SLACK_BOT_TOKEN
Get Slack App Token:
- Go to Basic information under Settings
- Navigate to App-Level Tokens and click Generate Token and Scopes
- Enter a token name, click Add Scope
- Add the scope "connections:write" and click Generate
- Copy the Token and save it in your .env file as SLACK_APP_TOKEN
Final Settings:
- Navigate to Socket Mode under Settings section.
- Toggle Enable Socket Mode on
- Navigate to Interactivity & Shortcuts under Features section.
- Toggle Interactivity on
- Navigate to Event Subscriptions under Features section.
- Toggle Enable Events on
- Click Subscribe to bot events>Add Bot User Event
- Add events "message.im" and "app_mention"
- Click *Save Changes at the bottom
- Click on Reinstall your app that appears on the banner
- Click Allow

Yup, that's it we are done!!

Get Tavily API keys

Sign Up or Log In, Get API Keys:
- Visit Tavily.
- If you don't have an account, sign up. Otherwise, log in with your existing credentials.
- Copy the API Key and save it in your .env as TAVILY_API_KEY

Environment Setup

Create a Virtual Environment: This isolates your Python setup on a per-project basis.
- Create a virtual environment using:
```
python -m venv venv
```
- Activate the virtual environment:
  - For Linux or macOS:
```
source venv/bin/activate
```
  - For Windows:
```
venv\Scripts\activate
```

Install Dependencies:

Install the required libraries. Run the command:

pip install requests langchain_openai langchain_core python-dotenv langchain_community langchainhub slack_bolt wikipedia

Environment Variables:

Set up variable in a .env file in the script's directory:

OPENAI_API_KEY=your_openai_api_key
SLACK_BOT_TOKEN=your_slack_bot_token
SLACK_APP_TOKEN=your_slack_app_token
TAVILY_API_KEY=your_serper_api_key

Get those agents to work

Run the app:
- Run the app:
```
python research_agent.py
```
Run agents on Slack
- Open your Slack app
- In the Apps section you should have your app
- Click on the app and provide it with the company's website you want to analyze
- Sit back and let the bot analyze

Notes

Handle API keys and configurations carefully to maintain script functionality.
Test changes in a controlled environment before deployment.

This README provides comprehensive instructions for customization and usage of the script. Enjoy coding!

ulfimlg/SlackResearchAgent