Name	Name	Last commit message	Last commit date
parent directory ..
README.md	README.md
agent-configuration.md	agent-configuration.md
javascript-actions.md	javascript-actions.md

AI Teammate Workflows

AI Teammate is a powerful automation pattern that combines DMTools, Cursor AI, and GitHub Actions to create intelligent development workflows.

What is AI Teammate?

AI Teammate is a workflow automation system that:

✅ Reads Jira tickets as input
✅ Uses AI (Cursor) to perform complex tasks
✅ Executes DMTools CLI commands for integrations
✅ Runs JavaScript for post-processing
✅ Creates PRs and updates tickets automatically

Architecture

┌─────────────────────────────────────────────────────────────────────┐
│                         GitHub Actions Workflow                      │
│  ┌───────────────────────────────────────────────────────────────┐  │
│  │                                                                │  │
│  │  1. Trigger → 2. Setup → 3. DMTools Run → 4. Post-Processing │  │
│  │                                                                │  │
│  └───────────────────────────────────────────────────────────────┘  │
└─────────────────────────────────────────────────────────────────────┘
                                    │
                                    ▼
┌─────────────────────────────────────────────────────────────────────┐
│                          Teammate Job                                │
│  ┌───────────────────────────────────────────────────────────────┐  │
│  │ - Fetch ticket from Jira                                       │  │
│  │ - Prepare input context (ticket + children + attachments)     │  │
│  │ - Execute CLI commands (run cursor-agent)                     │  │
│  │ - Run post-JS action                                          │  │
│  └───────────────────────────────────────────────────────────────┘  │
└─────────────────────────────────────────────────────────────────────┘
                                    │
                                    ▼
┌─────────────────────────────────────────────────────────────────────┐
│                          Cursor Agent                                │
│  ┌───────────────────────────────────────────────────────────────┐  │
│  │ - Read input folder (ticket context)                           │  │
│  │ - Perform AI task (write code, documentation, etc.)           │  │
│  │ - Use dmtools CLI for Jira/Confluence/Figma access           │  │
│  │ - Write output to outputs/response.md                         │  │
│  └───────────────────────────────────────────────────────────────┘  │
└─────────────────────────────────────────────────────────────────────┘
                                    │
                                    ▼
┌─────────────────────────────────────────────────────────────────────┐
│                        JavaScript Action                             │
│  ┌───────────────────────────────────────────────────────────────┐  │
│  │ - Read outputs/response.md                                     │  │
│  │ - Create git branch & commit                                   │  │
│  │ - Push changes & create PR                                     │  │
│  │ - Update Jira ticket (status, comments, labels)               │  │
│  └───────────────────────────────────────────────────────────────┘  │
└─────────────────────────────────────────────────────────────────────┘

Quick Start

Prerequisites

GitHub repository with DMTools configured
GitHub Actions enabled
Secrets configured (see Configuration)
Cursor API key for AI operations

Step 1: Set Up GitHub Actions Workflow

Create .github/workflows/ai-teammate.yml:

name: AI Teammate

on:
  workflow_dispatch:
    inputs:
      config_file:
        description: 'Path to config'
        required: true
      encoded_config:
        description: 'Encoded or JSON Agent Config'
        required: false

permissions:
  contents: write
  pull-requests: write
  actions: read
  
jobs:
  cursor-agent:
    runs-on: ubuntu-latest
    
    steps:
    - name: Checkout repository
      uses: actions/checkout@v4

    - name: Set up Java 23
      uses: actions/setup-java@v4
      with:
        distribution: 'temurin'
        java-version: '23'

    - name: Install Cursor CLI
      run: |
        curl https://cursor.com/install -fsS | bash
        echo "$HOME/.local/bin" >> $GITHUB_PATH

    - name: Install DMTools CLI
      run: |
        curl -fsSL https://github.com/IstiN/dmtools/releases/latest/download/install.sh | bash
        echo "$HOME/.dmtools/bin" >> $GITHUB_PATH

    - name: Run AI Teammate
      env:
        CURSOR_API_KEY: ${{ secrets.CURSOR_API_KEY }}
        JIRA_EMAIL: ${{ secrets.JIRA_EMAIL }}
        JIRA_API_TOKEN: ${{ secrets.JIRA_API_TOKEN }}
        JIRA_BASE_PATH: ${{ vars.JIRA_BASE_PATH }}
        GEMINI_API_KEY: ${{ secrets.GEMINI_API_KEY }}
        PAT_TOKEN: ${{ secrets.PAT_TOKEN }}
        DMTOOLS_INTEGRATIONS: "jira,confluence,figma,ai,cli,file"
      run: |
        dmtools run "${{ inputs.config_file }}" "${ENCODED_CONFIG:-}"

Step 2: Create Agent Configuration

Create agents/story_description.json:

{
  "name": "Teammate",
  "params": {
    "agentParams": {
      "aiRole": "Experienced Business Analyst",
      "instructions": [
        "Read ticket details from input folder",
        "Write enhanced story description following template",
        "Use dmtools CLI to read Confluence templates and Figma designs",
        "Check child tickets for questions and solution designs",
        "Write output to outputs/response.md"
      ],
      "formattingRules": "Use Jira markup format",
      "knownInfo": ""
    },
    "cliCommands": [
      "./cicd/scripts/run-cursor-agent.sh \"Enhance story description based on input\""
    ],
    "outputType": "field",
    "fieldName": "Description",
    "operationType": "Replace",
    "ticketContextDepth": 1,
    "skipAIProcessing": true,
    "inputJql": "key = PROJ-123",
    "postJSAction": "agents/assignForReview.js"
  }
}

Step 3: Trigger the Workflow

# Via GitHub CLI
gh workflow run ai-teammate.yml \
  -f config_file=agents/story_description.json

# Via GitHub UI
# Go to Actions → AI Teammate → Run workflow
# Enter: agents/story_description.json

Configuration

GitHub Secrets

Add these secrets in Settings → Secrets and variables → Actions → Secrets:

Secret	Description	Example
`CURSOR_API_KEY`	Cursor AI API key	`cur_...`
`JIRA_EMAIL`	Jira account email	`user@company.com`
`JIRA_API_TOKEN`	Jira API token	`ATATT3xFfGF0T...`
`GEMINI_API_KEY`	Google Gemini key (optional)	`AIza...`
`FIGMA_TOKEN`	Figma API token (optional)	`figd_...`
`PAT_TOKEN`	GitHub Personal Access Token for PR creation	`ghp_...`

GitHub Variables

Add these variables in Settings → Secrets and variables → Actions → Variables:

Variable	Description	Example
`JIRA_BASE_PATH`	Jira URL	`https://company.atlassian.net`
`JIRA_AUTH_TYPE`	Auth type	`Basic`
`CONFLUENCE_BASE_PATH`	Confluence URL	`https://company.atlassian.net/wiki`
`CONFLUENCE_DEFAULT_SPACE`	Default space	`TEAM`
`FIGMA_BASE_PATH`	Figma API URL	`https://api.figma.com/v1`

Agent Configuration Structure

Agent configurations follow this structure:

{
  "name": "Teammate",
  "params": {
    // Agent AI parameters
    "agentParams": {
      "aiRole": "Role description",
      "instructions": ["Instruction 1", "Instruction 2"],
      "formattingRules": "Output format rules",
      "knownInfo": "Additional context",
      "fewShots": "Example outputs"
    },
    
    // CLI commands to execute
    "cliCommands": [
      "./scripts/run-cursor-agent.sh \"Task description\""
    ],
    
    // Output configuration
    "outputType": "field|comment|creation|none",
    "fieldName": "Description",
    "operationType": "Replace|Append",
    
    // Ticket selection
    "inputJql": "key = PROJ-123",
    "ticketContextDepth": 0,
    
    // Processing options
    "skipAIProcessing": true|false,
    "attachResponseAsFile": true|false,
    
    // Post-processing
    "postJSAction": "agents/myAction.js",
    
    // User context
    "initiator": "user-account-id"
  }
}

Configuration Options

`agentParams`

AI agent configuration:

aiRole: Role/persona for the AI (e.g., "Senior Java Developer")
instructions: Array of instructions for the AI
formattingRules: Output format requirements
knownInfo: Additional context or knowledge
fewShots: Example outputs for few-shot learning

`cliCommands`

Commands to execute:

"cliCommands": [
  "./cicd/scripts/run-cursor-agent.sh \"Implement feature X\""
]

`outputType`

Where to put results:

field: Update a Jira field
comment: Add as Jira comment
confluence: Create/update Confluence page
none: No automatic output

`ticketContextDepth`

How much context to include:

0: Only the ticket itself
1: Ticket + direct children (subtasks)
2: Ticket + children + linked tickets

`skipAIProcessing`

true: Use CLI output directly (no AI processing)
false: Process through AI (default)

`postJSAction`

JavaScript file to run after completion:

"postJSAction": "agents/developTicketAndCreatePR.js"

Use Cases & Examples

Use Case 1: Story Description Enhancement

Goal: Enhance story descriptions with AI

Configuration: agents/story_description.json

{
  "name": "Teammate",
  "params": {
    "agentParams": {
      "aiRole": "Senior Business Analyst",
      "instructions": [
        "Read ticket from input folder",
        "Enhance description following Confluence template",
        "Use dmtools to read child tickets for context",
        "Check Figma designs if referenced",
        "Write enhanced description to outputs/response.md"
      ],
      "formattingRules": "Use Jira markup: h2., *, {code}, etc."
    },
    "cliCommands": [
      "./cicd/scripts/run-cursor-agent.sh \"Enhance story description\""
    ],
    "outputType": "field",
    "fieldName": "Description",
    "operationType": "Replace",
    "ticketContextDepth": 1,
    "skipAIProcessing": true,
    "inputJql": "key = DMC-532",
    "postJSAction": "agents/assignForReview.js"
  }
}

Workflow:

Fetches ticket DMC-532
Includes child tickets (questions, solution design)
Cursor agent enhances description
Updates ticket description
Assigns to initiator for review

Use Case 2: Automated Development

Goal: Implement ticket and create PR

Configuration: agents/story_development.json

{
  "name": "Teammate",
  "params": {
    "agentParams": {
      "aiRole": "Senior Java Software Engineer",
      "instructions": [
        "Read ticket requirements from input folder",
        "Implement code changes following architecture",
        "Create unit tests with good coverage",
        "Write development summary to outputs/response.md",
        "DO NOT create branches or push - handled by post-JS action"
      ],
      "formattingRules": "Markdown with sections: Approach, Files Modified, Test Coverage, Issues/Notes"
    },
    "cliCommands": [
      "./cicd/scripts/run-cursor-agent.sh \"Implement ticket\""
    ],
    "outputType": "none",
    "ticketContextDepth": 1,
    "skipAIProcessing": true,
    "postJSAction": "agents/developTicketAndCreatePR.js"
  }
}

Workflow:

Fetches ticket with full context
Cursor agent implements code + tests
Writes summary to outputs/response.md
JS action creates branch, commits, pushes
Creates PR with summary as description
Updates ticket status to "In Review"
Posts comment with PR link

Use Case 3: Solution Design Creation

Goal: Create solution design tickets

Configuration: agents/createSolutionDesignTicketsAndAssignForReview.js

{
  "name": "Teammate",
  "params": {
    "agentParams": {
      "aiRole": "Solution Architect",
      "instructions": [
        "Analyze story requirements",
        "Create solution design document",
        "Break down into technical tasks",
        "Identify risks and dependencies"
      ]
    },
    "cliCommands": [
      "./cicd/scripts/run-cursor-agent.sh \"Create solution design\""
    ],
    "outputType": "comment",
    "ticketContextDepth": 1,
    "postJSAction": "agents/createSolutionDesignTicketsAndAssignForReview.js"
  }
}

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

README.md

AI Teammate Workflows

What is AI Teammate?

Architecture

Quick Start

Prerequisites

Step 1: Set Up GitHub Actions Workflow

Step 2: Create Agent Configuration

Step 3: Trigger the Workflow

Configuration

GitHub Secrets

GitHub Variables

Agent Configuration Structure

Configuration Options

`agentParams`

`cliCommands`

`outputType`

`ticketContextDepth`

`skipAIProcessing`

`postJSAction`

Use Cases & Examples

Use Case 1: Story Description Enhancement

Use Case 2: Automated Development

Use Case 3: Solution Design Creation

See Also

FilesExpand file tree

ai-teammate

Directory actions

More options

Directory actions

More options

Latest commit

History

ai-teammate

Folders and files

parent directory

README.md

AI Teammate Workflows

What is AI Teammate?

Architecture

Quick Start

Prerequisites

Step 1: Set Up GitHub Actions Workflow

Step 2: Create Agent Configuration

Step 3: Trigger the Workflow

Configuration

GitHub Secrets

GitHub Variables

Agent Configuration Structure

Configuration Options

agentParams

cliCommands

outputType

ticketContextDepth

skipAIProcessing

postJSAction

Use Cases & Examples

Use Case 1: Story Description Enhancement

Use Case 2: Automated Development

Use Case 3: Solution Design Creation

See Also

`agentParams`

`cliCommands`

`outputType`

`ticketContextDepth`

`skipAIProcessing`

`postJSAction`