GitHub Agentic Workflows GitHub Agentic Workflows
GitHub Agentic Workflows (gh-aw) lets developers define AI-powered repository automation in Markdown with YAML frontmatter and compile it into GitHub Actions workflows. A single `gh aw` command turns a natural-language workflow into a locked-down `.lock.yml` that runs coding agents on issues, pull requests, schedules, or commands.
The project reached v0.88.8 on September 8, 2026 with frequent weekly releases, crossed 5,100+ GitHub stars, and continues to add engines and safe-output types; it is developed in the open by GitHub Next.
As of
- Markdown + YAML frontmatter workflow definitions compiled to GitHub Actions
- Pluggable engines: GitHub Copilot, Claude Code, OpenAI Codex, Google Gemini, and Pi
- Safe-outputs model: agents propose issues, comments, PRs, and labels that are applied through sandboxed steps
- MCP tool configuration with a curated GitHub MCP server and custom servers
- Network sandboxing, permission scoping, and automatic secret redaction
- Triggers on issues, PRs, discussions, schedules, and slash commands
- Shared reusable workflows and a `gh aw` CLI for compile, run, logs, and audit
Maintainers and platform teams who want to run coding agents on repository events inside GitHub Actions with auditable, version-controlled configuration and strict output controls.
Teams not hosted on GitHub, projects that cannot spend Actions minutes and model tokens on automation, or users looking for an interactive local coding agent rather than event-driven repository automation.
The gh-aw CLI extension and compiler are free and MIT-licensed. Running workflows consumes GitHub Actions minutes and the tokens of the chosen AI engine (GitHub Copilot, Anthropic, OpenAI, Google), which are billed by those providers.
GitHub Agentic Workflows is the most GitHub-native way to run agents on repository events, with a security posture (safe-outputs, sandboxing) that is stronger than hand-rolled Actions; it is still pre-1.0 and evolving quickly.
Is GitHub Agentic Workflows free?
Yes - GitHub Agentic Workflows is Open Source. The gh-aw CLI extension and compiler are free and MIT-licensed. Running workflows consumes GitHub Actions minutes and the tokens of the chosen AI engine (GitHub Copilot, Anthropic, OpenAI, Google), which are billed by those providers.
Is GitHub Agentic Workflows open source?
Yes - GitHub Agentic Workflows is open source. The gh-aw CLI extension and compiler are free and MIT-licensed. Running workflows consumes GitHub Actions minutes and the tokens of the chosen AI engine (GitHub Copilot, Anthropic, OpenAI, Google), which are billed by those providers.
Who is GitHub Agentic Workflows best for?
Maintainers and platform teams who want to run coding agents on repository events inside GitHub Actions with auditable, version-controlled configuration and strict output controls.
Who is GitHub Agentic Workflows not ideal for?
Teams not hosted on GitHub, projects that cannot spend Actions minutes and model tokens on automation, or users looking for an interactive local coding agent rather than event-driven repository automation.
What are the best GitHub Agentic Workflows alternatives?
The closest GitHub Agentic Workflows alternatives on ai.dosa.dev are Jules, OpenHands, SWE-agent - all listed under Autonomous & Async Agents.
Who makes GitHub Agentic Workflows?
GitHub Agentic Workflows is developed by GitHub. It is listed in the Autonomous & Async Agents category on ai.dosa.dev.
Favorite this tool to revisit it later, or Zap it to contribute to the public vote count.
Content on this page is AI-generated. Please verify details with the vendor's website for accuracy.