GitHub Agentic Workflows, or gh-aw, lets teams describe repository automation in Markdown and compile it into hardened GitHub Actions workflows that run AI coding agents. The official Public Preview supports GitHub Copilot, Claude Code, OpenAI Codex, Google Gemini and custom processors for scheduled or event-triggered work such as issue triage, pull-request review and repository maintenance. Its guardrail model uses sandboxed execution, read-only defaults, scoped permissions, network controls and gated safe outputs so agents do not receive unrestricted write access. The project also exposes per-run budgets, spend visibility and OpenTelemetry-based cost analysis. Buyers should still treat it as a fast-moving preview, review generated lock files and granted permissions, and account for prompt-injection risk plus the data policies and usage charges of the selected model provider. It is best suited to GitHub-centric teams that want agentic automation to remain inside their existing Actions governance.
GitHub Agentic Workflows
Run AI coding agents in GitHub Actions from Markdown
Open-source GitHub CLI extension for defining repository automations in Markdown and running Copilot, Claude Code, Codex, Gemini or custom agents inside hardened GitHub Actions workflows.
Pricing
Free and open source under the MIT license. Operating cost comes from GitHub Actions usage and the selected AI engine or model provider; no separate gh-aw license fee is documented.
Platforms
GitHub CLI extension and GitHub Actions workflow compiler. Markdown source files live in .github/workflows and compile to locked Actions YAML; requires Actions plus credentials for the chosen AI engine.
Categories
Tags
Alternatives
Claude Code GitHub Action
Anthropic's official GitHub Action that runs Claude Code in CI
Anthropic's official GitHub Action that brings Claude Code into GitHub PRs and issues — answering questions, reviewing code, and implementing changes when triggered by @claude mentions or workflow automation. This is the CI/CD Action, not the terminal agent.
OpenAI Codex GitHub Action
OpenAI's official GitHub Action that runs Codex in CI
OpenAI's official GitHub Action for running Codex CLI in GitHub Actions workflows with filesystem/network permission profiles and runner-level privilege controls. This is the CI/CD Action, not the terminal Codex agent; current Codex exec enables analytics by default, so telemetry policy needs review.
Related Tools
MEDUSA
AI-first security scanner for LLM, agent, MCP, and RAG codebases
MEDUSA is an AGPL-3.0 AI-first security scanner from Pantheon Security that checks AI and machine-learning applications, LLM agents, MCP workflows, RAG pipelines, repository-poisoning risks, secrets, and agent-specific compromise patterns.
Mergify
Merge queue, CI insights, flaky-test controls, and stacked pull requests for GitHub teams
Mergify is a pull request automation platform that keeps main branches green with merge queue batching, merge protections, CI Insights, flaky-test detection, and stacked pull requests. Its Stacks workflow turns commits on one local branch into focused PR chains, helping teams review large AI-generated or feature-heavy changes without losing queue safety.
Terrateam
Open-source GitOps automation for Terraform and OpenTofu pull requests
Terrateam is open-source GitOps infrastructure orchestration for Terraform and OpenTofu pull requests. It automates plans and applies in GitHub workflows, supports monorepos and many workspaces, and adds apply-only locks, OPA/Rego policy checks, cost and drift signals, and approval controls without forcing teams into a separate IaC platform.
Digger
Open-source IaC orchestration that runs Terraform plans and applies from pull request comments
Digger is an open-source infrastructure-as-code orchestration tool for running Terraform and OpenTofu plan/apply workflows from pull request comments. It uses the team’s existing VCS and CI runners instead of adding a separate runner fleet, supports GitHub, GitLab, and Azure DevOps style workflows, and stores PR-level locks and plan cache in the user’s cloud account.
Twill AI
Autonomous coding agents that ship while you sleep
Twill is an autonomous coding agent platform that implements features, fixes bugs, and ships pull requests without manual intervention. Uses structured workflow of research, planning, human review, implementation in isolated sandbox, AI code review, then merge. Supports custom agent configurations with multiple LLM providers, isolated dev environments for verification, and integrations with GitHub, Linear, Sentry, Notion, and cloud platforms for end-to-end engineering automation.
Poethepoet
Task runner for Python with Poetry and uv
Poethepoet (poe) is a batteries-included task runner for Python projects that integrates with Poetry and uv package managers. Define tasks in pyproject.toml, compose them in sequential, parallel, or DAG workflows, and execute with full virtual environment context. Supports shell commands, Python scripts, environment variables, .env file loading, and auto-generated shell completion across bash, zsh, and fish for streamlined development workflows.