New ways to automate, govern, and scale AI across GitHub

[GitHub]

The latest launches from GitHub. See what’s new and how to get started.

‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 

JULY EDITION New agentic experiences, budget controls, and governance options

TOP RELEASE TO TRY Work with agents how and where you want With GitHub Agentic Workflows, you can automate reasoning-based tasks like issue triage, CI failure analysis, and documentation updates by leveraging coding agents inside GitHub Actions. GitHub Agentic Workflows is now in public preview. Useful if you want to: Define automations in natural language Have agents tackle triage and failure analysis Automate documentation updates Learn more about GitHub Agentic Workflows

COPILOT GitHub Copilot app is now generally available Utilize agent-driven development directly on your desktop in an environment built natively on GitHub. Try out the GitHub Copilot app

COPILOT SDK Build an agent into any app with Copilot SDK Embed GitHub Copilot's agentic engine into your applications, services, and tools with direct access through a stable API available in multiple languages. GitHub SDK is now generally available. Discover Copilot SDK

ENTERPRISE TEAMS Organize your teams at enterprise scale Group users into teams that can bridge across organizations within your enterprise. Enterprise teams are now generally available. Use enterprise teams to organize your users

More updates this month

Manage and track budgets through API Programmatically manage your budgets and fetch usage summaries without needing to access the UI. Track AI credit consumption per user The Copilot usage metrics API now includes AI credits each user consumed per day, derived from the usage-based billing API. Set hard budget limits in GitHub Advanced Security (GHAS) Place hard budget limits to block additional license usage once the threshold is reached. Run Copilot in dedicated sandboxes Use either local or cloud sandboxes to work with Copilot in a secure environment. Copilot code review is easier to customize to your needs With the latest updates to Copilot code review, you can set the runner type, include larger custom instruction files, and configure content exclusion settings. Customize your Copilot code review needs around your team Configure Copilot code review to use agent skills and MCP server connections that provide your organization’s context. Set review effort levels per repository to match reasoning needs. Control which models organizations can access As new models become available in GitHub Copilot—including Claude Sonnet 5, Claude Fable 5, Claude Opus 4.8 Fast Mode, and MAI Code 1 Flash—use model rules to manage access at the organization level and apply more granular controls than enterprise-wide policies alone.

Was this monthly product update helpful? Very helpful! Somewhat helpful Not relevant to me

See you next month, Mona and the team at GitHub

‌  ‌  ‌  ‌  Sign in to GitHub  •  Unsubscribe Privacy policy  •  Terms of use  •  Contact us GitHub, Inc.  •  88⁠ Colin P⁠ Kelly Jr Street  •  San⁠ Francisco,⁠ CA⁠ 94107 You are receiving this because you are the administrator of a GitHub account.