GitHub Connector

Configure the GitHub connector to enable AI Team members to review pull requests, manage issues, and interact with repositories.

Overview

The GitHub connector enables AI Team members to interact with your GitHub repositories, providing capabilities for code review, issue management, pull request operations, and workflow monitoring. By connecting GitHub to AI Team, AI teammates can review code changes, manage issues, monitor CI/CD workflows, and help coordinate development activities.

The connector provides access to pull requests, issues, repository files, GitHub Actions workflows, and security scanning results, allowing AI teammates to assist with code quality analysis, issue triage, and development workflow troubleshooting.

Add the GitHub Connector

To add the GitHub connector, you authenticate with GitHub using either OAuth or a Personal Access Token.

Prerequisites

Before configuring the connector, ensure you have:

  • GitHub account with access to the repositories you need
  • Appropriate repository permissions based on your use cases
  • Ability to generate a Personal Access Token or complete OAuth authorization

Configuration Steps

  1. Navigate to AI Team > Connectors in the Edge Delta application
  2. Find the GitHub connector
  3. Click the connector card to open the configuration panel
  4. Choose authentication method:
    • OAuth: Click Connect with OAuth and authorize in GitHub
    • Personal Access Token: Enter your GitHub PAT
  5. Click Save

When using OAuth, after authentication you’ll see the connected state:

The connector is now available for use by AI Team members who have been assigned this connector.

General Options

GitHub PAT

Personal Access Token for GitHub authentication. Use this as an alternative to OAuth. To create a PAT, go to GitHub → Settings → Developer settings → Personal access tokens and generate a new token (classic) with appropriate scopes: repo for repository access, workflow for GitHub Actions access, and read:org for organization data. See Managing your personal access tokens for detailed instructions.

Tools

The GitHub connector provides over 80 tools for interacting with repositories, pull requests, issues, workflows, and security features. Key tools include:

Pull Request Operations

  • get_pull_request - Get pull request details
  • get_pull_request_diff - Get the diff of a pull request
  • get_pull_request_files - Get files changed in a PR
  • list_pull_requests - List pull requests in a repository
  • create_pull_request - Create a new pull request
  • update_pull_request - Update an existing pull request
  • merge_pull_request - Merge a pull request
  • request_copilot_review - Request GitHub Copilot code review for a PR

Pull Request Reviews

  • create_pending_pull_request_review - Create a pending review
  • add_comment_to_pending_review - Add review comments
  • submit_pending_pull_request_review - Submit the pending review
  • get_pull_request_reviews - Get reviews for a PR
  • get_pull_request_review_comments - Get review comments

Issue Management

  • get_issue - Get issue details
  • list_issues - List issues in a repository
  • create_issue - Create a new issue
  • update_issue - Update an existing issue
  • add_issue_comment - Add a comment to an issue
  • add_sub_issue - Add a sub-issue to a parent issue
  • list_sub_issues - List sub-issues for an issue

Repository Operations

  • get_file_contents - Get contents of a file or directory
  • create_or_update_file - Create or update a file
  • delete_file - Delete a file
  • list_branches - List branches in a repository
  • create_branch - Create a new branch
  • list_commits - Get list of commits
  • get_commit - Get commit details

Workflow and CI/CD

  • list_workflows - List workflows in a repository
  • list_workflow_runs - List workflow runs
  • get_workflow_run - Get workflow run details
  • list_workflow_jobs - List jobs for a workflow run
  • get_job_logs - Download logs for a workflow job
  • summarize_run_log_failures - Analyze why a workflow run failed
  • summarize_job_log_failures - Summarize failed job logs
  • rerun_workflow_run - Re-run an entire workflow
  • rerun_failed_jobs - Re-run only failed jobs

Security Scanning

  • list_code_scanning_alerts - List code scanning alerts
  • get_code_scanning_alert - Get code scanning alert details
  • list_dependabot_alerts - List Dependabot alerts
  • get_dependabot_alert - Get Dependabot alert details
  • list_secret_scanning_alerts - List secret scanning alerts
  • get_secret_scanning_alert - Get secret scanning alert details

Search Operations

  • search_code - Search code across repositories
  • search_issues - Search for issues
  • search_pull_requests - Search for pull requests
  • search_repositories - Find repositories
  • search_users - Find users

GitHub Copilot

  • assign_copilot_to_issue - Assign Copilot to work on an issue
  • create_pull_request_with_copilot - Delegate task to Copilot coding agent
  • get_copilot_space - Get additional context from a Copilot space
  • list_copilot_spaces - List accessible Copilot spaces

Projects

  • list_projects - List projects for a user or organization
  • get_project - Get project details
  • list_project_items - List project items
  • add_project_item - Add item to a project

How to Use the GitHub Connector

The GitHub connector integrates with AI Team, enabling AI teammates to review code, manage issues, and troubleshoot workflows based on natural language queries. Once configured, AI teammates can interact with your repositories to help with development workflows.

Use Case: Pull Request Review

When a developer opens a pull request, AI teammates can review the changes for potential issues, security vulnerabilities, or code quality concerns. For example, when asked “Review PR #123”, the AI can examine the diff, identify potential problems, and provide feedback before human reviewers spend time on the PR.

Use Case: Issue Management

AI teammates can help create, update, and organize GitHub issues. When investigating incidents or planning work, the AI can create issues with appropriate details, add comments with findings, and link related issues together. This helps keep development work organized and tracked.

Use Case: Workflow Debugging

When GitHub Actions workflows fail, AI teammates can analyze the failure, examine logs, and identify the root cause. For example, when asked “Why did the build fail?”, the AI can check recent workflow runs, review error logs, and suggest specific fixes based on the failure patterns.

Troubleshooting

Authentication errors: Verify your Personal Access Token is valid and hasn’t expired. Ensure it has the required scopes (repo, workflow, read:org). For OAuth, try disconnecting and reconnecting to reauthorize.

Rate limit errors: GitHub enforces API rate limits (5,000 requests per hour for authenticated requests). If you hit this limit, wait for it to reset or reduce the frequency of requests.

Repository access errors: Verify the authenticated account has access to the repository. For private repositories, ensure the PAT has the full repo scope.

Permission errors: Ensure your token has write permissions for operations like merging PRs or creating issues. Some operations require specific scopes or repository permissions.

Next Steps

For additional help, visit AI Team Support.