SREBot

A Site Reliability Engineering (SRE) bot that integrates with Slack, GitHub, and Checkly to help manage incidents and improve Mean Time To Resolution (MTTR).

Features

• Real-time incident management through Slack
• Integration with Checkly for monitoring alerts
• GitHub integration for tracking releases and changes
• AI-powered assistance using OpenAI
• Automatic context aggregation for incidents

Prerequisites

• Node.js v22.11.0
• Docker and Docker Compose
• PostgreSQL (via Docker or local installation)
• Accounts and API keys for:
  • OpenAI
  • Slack
  • GitHub
  • Checkly

Setup

1. Environment Variables

Duplicate the env.example file in the root directory and add your keys as per requirements

2. Database Setup

The project uses PostgreSQL as its database. To set it up using Docker:

1. Start the PostgreSQL container:

   docker compose up -d
   

2. Run database migrations:

   npm run db:migrate
   

3. (Optional) To explore the database using Prisma Studio:

   npm run db:studio
   

To reset the database if needed:

docker compose down -v  # Remove containers and volumes
docker compose up -d    # Start fresh
npm run db:migrate     # Run migrations again

3. Installation

1. Install dependencies

npm install

1. Initialize the OpenAI assistant

   • First, get your OpenAI API key
   • Add the API key to your .env file
   • Create the assistant by running:

   npx ts-node scripts/init-assistant.ts
   

Go to the OpenAI Portal to find your assistant ID

Running the Application

There are several ways to run the application depending on your needs:

Development Mode

npm run dev

Slack Bot Only

npm run bot:start

Production Mode

npm run build
npm start

Running Tests

npm test

Project Structure

• /src/aggregator: Context aggregation logic for incidents
• /src/ai: AI assistant and tool implementations
• /src/checkly: Checkly API integration
• /src/github: GitHub API integration
• /src/routes: API routes for webhooks
• /src/slackbot: Slack bot implementation
• /src/sre-assistant: Main SRE assistant logic
• /prisma: Database schema and migrations

Available Commands

• npm run dev: Start development server
• npm run bot:start: Start Slack bot only
• npm run build: Build the application
• npm start: Start the built application
• npm test: Run tests
• npm run db:migrate: Run database migrations
• npm run db:studio: Open Prisma Studio
• npm run db:deploy: Deploy database migrations
• npm run db:generate: Generate Prisma client

External Service Setup

Slack Setup

1. Create a new Slack app in your workspace
2. Configure Bot Token Scopes:
   • chat:write
   • app_mentions:read
   • commands
3. Install the app to your workspace
4. Copy the signing secret, bot token, and app token to your .env file

GitHub Setup

1. Create a Personal Access Token with repo permissions
2. Configure webhook in your organization/repository:
   • Payload URL: your-server/github-webhook
   • Content type: application/json
   • Secret: Same as GH_WEBHOOK_SECRET in .env
   • Events: Release events

Checkly Setup

1. Get your API key and Account ID from Checkly
2. Configure webhook in Checkly:
   • URL: your-server/checkly-webhook
   • Select relevant alert types

Architecture

sequenceDiagram
    participant CLY as Checkly
    participant APP as App
    participant GHB as Github
    participant NOT as Notion API
    participant SLCK_API as Slack API
    participant CLY_API as Checkly Api
    participant AI as AI Model(OpenAI)
    participant SLCK as Slack
    CLY->>APP:Alert Webhook
    critical Aggregate Context
      APP->>GHB:Get Releases
      GHB->>AI:Identify Relevant Releases
      GHB->>AI:Identify Relevant Deployments
      APP->>NOT:Get Knowledge Bases
      APP->>SLCK_API:Get Channel Summary
      SLCK_API->>AI:Summarize Channel
      APP->>CLY_API:Get Alert Details
    end

    APP->>AI:Ananlyze context, Root Cause and Create Summary
    AI->>APP:Alert Summary
    APP->>SLCK:Send Alert Message

License

ISC License - see the LICENSE file for details