Contributing to Frontier

September 23, 2026 ยท View on GitHub

Thank you for contributing! This guide ensures all contributors follow the same workflow, with or without GitHub Copilot.

Core Workflow (Mandatory)

The Issue-First Rule

ALL work MUST start with a GitHub Issue created BEFORE you write any code.

1. Create Issue -> 2. Claim Issue -> 3. Write Code -> 4. Commit with Issue# -> 5. Close Issue

Why? This ensures:

  • Work is tracked and visible
  • No duplicate efforts
  • Clear audit trail
  • Agent coordination works

Quick Start

Script-Based Repo Setup

Use the script-based flow when you are setting up a local clone of the Frontier repository for contribution work.

# Clone the repository
git clone https://github.com/jnPiyush/AgentX.git
cd AgentX

# Run setup script
.\install.ps1

# Clone the repository
git clone https://github.com/jnPiyush/AgentX.git
cd AgentX

# Run setup script
./install.sh

Default install mode is local. GitHub and Azure DevOps modes are available for remote workflow integration.

If Frontier detects Azure-oriented files such as azure.yaml, .azure/, Azure Functions config, or Bicep, it can also recommend or install the Azure MCP Extension so the Azure Skills plugin is available when the app targets Azure.

First Time Setup

# Install GitHub CLI (if not already installed)
# Windows: winget install GitHub.cli
# Mac: brew install gh
# Linux: See https://github.com/cli/cli#installation

# Authenticate with GitHub
gh auth login

For Every Task

# 1. CREATE issue (choose the right template)
gh issue create --web
# OR use CLI:
gh issue create --title "[Type] Description" --label "type:story"

# 2. CLAIM the issue (move to 'In Progress' in Projects board)
# Use GitHub Projects UI to drag issue to 'In Progress' column

# 3. Create a branch
git checkout -b issue-<ISSUE#>-description

# 4. Make your changes
# (Code, test, document)

# 5. Commit with issue reference
git commit -m "type: description (#<ISSUE#>)"

# 6. Push and create PR
git push origin issue-<ISSUE#>-description
gh pr create --fill

# 7. After merge, close issue
gh issue close <ISSUE#> --comment "Completed in commit <SHA>"

Issue Types

TypeWhen to UseExample
EpicLarge initiative, multiple features"Build user authentication system"
FeatureNew capability"Add OAuth login"
StorySmall, specific task"Add logout button"
BugSomething broken"Login page returns 500"
SpikeResearch/investigation"Evaluate PostgreSQL vs MongoDB"
DocsDocumentation only"Update README with setup steps"

Creating Issues

Via Web (Easiest):

gh issue create --web

This opens GitHub in your browser with structured templates.

Via CLI (Faster):

# Story
gh issue create --title "[Story] Add logout button" --label "type:story"

# Bug
gh issue create --title "[Bug] Login returns 500" --label "type:bug,priority:p0"

# Feature (has UI work)
gh issue create --title "[Feature] Add OAuth login" --label "type:feature,needs:ux"

Commit Message Format

type: description (#issue-number)

Optional longer explanation

Types

  • feat - New feature
  • fix - Bug fix
  • docs - Documentation only
  • test - Adding/updating tests
  • refactor - Code restructuring
  • perf - Performance improvement
  • chore - Maintenance tasks

Examples

git commit -m "feat: add OAuth login support (#123)"
git commit -m "fix: resolve 500 error on login page (#124)"
git commit -m "docs: update README with setup instructions (#125)"

Testing Requirements

All code must have tests with >=80% coverage

# Run tests
dotnet test                    # C#
pytest                         # Python
npm test                       # JavaScript

# Check coverage
dotnet test /p:CollectCoverage=true
pytest --cov=src --cov-report=html
npm run test:coverage

Test Pyramid

  • 70% Unit tests (fast, isolated)
  • 20% Integration tests (components working together)
  • 10% E2E tests (full user flows)

Security Checklist

Before committing:

  • No secrets/credentials in code
  • Environment variables used for sensitive data
  • SQL queries use parameterization (NEVER concatenation)
  • Input validation implemented
  • Dependencies scanned for vulnerabilities
# Scan for secrets (pre-commit hook does this)
git diff --staged | grep -E '(password|api[_-]?key|secret|token)' -i

# Check dependencies
dotnet list package --vulnerable
pip-audit                      # Python
npm audit                      # Node.js

Code Standards

Read Before Coding

  1. AGENTS.md - Workflow and behavior guidelines
  2. Skills.md - Technical standards index
  3. Relevant skill docs in .github/skills/ folder

Key Standards

Documentation

  • XML docs for all public APIs (C#)
  • Docstrings for all functions (Python)
  • JSDoc for exported functions (JavaScript)
  • README in each module folder

Code Review Process

  1. Self-Review: Check your own code before requesting review
  2. Automated Checks: CI must pass (tests, linting, security scan)
  3. Human Review: At least 1 approving review required
  4. Address Feedback: Respond to all comments
  5. Merge: Squash and merge after approval

Review Checklist for Reviewers

  • Code follows style guidelines
  • Tests are comprehensive
  • Documentation is clear
  • Security concerns addressed
  • Performance considerations noted

Development Tools

Required

  • Git - Version control
  • GitHub CLI (gh) - Issue/PR management
  • .NET 10+ (for C# projects)
  • Python 3.11+ (for Python projects)
  • PostgreSQL 16+ (for database projects)
  • VS Code with extensions:
    • GitHub Copilot (for AI assistance)
    • C# Dev Kit (for .NET)
    • Python Extension
    • PostgreSQL Extension
  • Docker Desktop (for containerized testing)

Common Mistakes to Avoid

DON'T

  • Start coding before creating an issue
  • Create issues retroactively after work is done
  • Commit without issue reference in message
  • Push directly to main/master
  • Hardcode secrets or sensitive data
  • Skip tests
  • Merge your own PRs without review

DO

  • Create issue FIRST, then code
  • Use descriptive issue titles
  • Reference issues in commits (#123)
  • Create feature branches
  • Use environment variables for config
  • Write tests as you code
  • Request reviews from teammates

Getting Help

Working with AI Agents

If you have GitHub Copilot, it will automatically:

  • Guide you through the Issue-First workflow
  • Suggest code following our standards
  • Help with testing and documentation
  • Enforce security best practices

Without Copilot, follow this guide manually and use:

  • GitHub CLI for issue management
  • Pre-commit hooks for validation
  • The provided templates and checklists

Learning Resources

License

By contributing, you agree that your contributions will be licensed under the project's license (see LICENSE).

Questions? Open an issue with the type:docs label and we'll help!