📖

Resources

Official docs, blog series, community links, and further reading

8+
Official Docs
4
Blog Posts
5
Repositories
19
Blog Series Parts
5
Community Articles
45+
Total Resources
✅

Getting Started Checklist

Get up and running with GitHub Agentic Workflows.

📚

Official Documentation

8 pages

The primary documentation site at github.github.io/gh-aw — your authoritative reference for everything gh-aw.

📘 Documentation Home

Essential
Main documentation site. Complete reference for concepts, setup, security, and workflow authoring.
github.github.io/gh-aw/

🚀 Quick Start Guide

Essential
Step-by-step guide: install gh-aw, add a starter workflow, compile, and run. The fastest path from zero to a working workflow.
gh-aw/setup/quick-start/

🛠️ Creating Agentic Workflows

Essential
How to write, compile, and deploy your first workflow. Covers gh aw init, gh aw compile, and gh aw run.
gh-aw/setup/creating-workflows/

💡 About Workflows (Overview)

Recommended
Conceptual overview: coding agents + tools in GitHub Actions. Explains frontmatter, safe-outputs, and the compilation model.
gh-aw/introduction/overview/

📝 Markdown Reference

Recommended
Detailed guide on writing effective workflow instructions. Covers templating, substitutions, content organization, and common pitfalls.
gh-aw/reference/markdown/

🔒 Security Architecture

Essential
Comprehensive security model: sandboxed execution, network isolation, SHA-pinned dependencies, tool allow-listing, compile-time validation.
gh-aw/security/

🔗 Official Shortlink

Supplementary
Official shortlink redirecting to the documentation site. Easy to remember and share: gh.io/gh-aw
gh.io/gh-aw

🤖 llms.txt (Agent Reference)

Supplementary
Machine-readable reference for AI agents working with gh-aw. Feed this to your coding agent for optimal gh-aw authoring.
github.com/github/gh-aw/blob/main/llms.txt
📰

GitHub Blog Posts

4 posts

Official announcements, changelog entries, and deep dives from the GitHub Blog.

Automate Repository Tasks with GitHub Agentic Workflows

Essential
The main announcement post (8-minute read). Introduces Continuous AI, the security model (read-only by default, safe-outputs), multi-engine support (Copilot CLI, Claude Code, OpenAI Codex), and billing.
✍️ Don Syme, Peli de Halleux📅 Feb 13, 2026

GitHub Agentic Workflows — Now in Technical Preview

Recommended
Concise changelog entry covering key benefits: natural language workflows, security-first design, easy setup, agentic authoring. Links to Peli’s Agent Factory (50+ specialized workflows).
✍️ GitHub Staff📅 Feb 13, 2026Changelog

Continuous AI in Practice: What Developers Can Automate Today

Essential
Explores the Continuous AI pattern: natural-language rules + agentic reasoning, executed continuously inside repositories. Covers triage, documentation, code quality, and status reports.
✍️ GitHub Staff📅 Feb 5, 2026

Copilot Code Review Now Runs on an Agentic Architecture

Recommended
Code review now uses agentic tool-calling for broader context gathering. Higher-quality findings, lower noise, more actionable guidance. Runs on GitHub Actions infrastructure.
✍️ GitHub Staff📅 Mar 5, 2026Changelog
🗂️

GitHub Repositories

5 repos

Open-source repositories powering GitHub Agentic Workflows — from the CLI extension to the firewall and sample packs.

github/gh-aw

Essential
The main repository. GitHub CLI extension for writing agentic workflows in natural language Markdown. MIT License.
⭐ 4,000+Go (70.8%), JS (27.6%)📦 v0.64.0313 releases

githubnext/agentics

Recommended
Sample pack of GitHub Agentic Workflows. Ready-to-use templates: daily-repo-status, issue-triage, and more. Great starting point.
📄 MarkdownTemplates & Samples

github/gh-aw-firewall

Recommended
Agent Workflow Firewall (AWF). Network egress control for AI agents with domain-based access control. Critical for security.
🔒 SecurityNetwork Isolation

gh-aw-actions

Supplementary
Shared library of custom GitHub Actions used by compiled agentic workflows. The building blocks behind the generated YAML.
⚙️ GitHub ActionsUnder github org

MCP Gateway

Supplementary
Routes Model Context Protocol (MCP) server calls through a unified HTTP gateway for agentic workflows. Companion project.
🔌 MCP ProtocolCompanion project
✍️

Community Articles

5 articles

Hands-on guides, deep dives, and practical perspectives from the developer community.

GitHub Agentic Workflows: A Hands-On Guide to AI-Powered CI/CD

Recommended
Built entire demo using Copilot CLI. Covers four workflow patterns: issue triage, PR review, docs sync, CI failure investigation. Honest assessment of rough edges. “Set up four workflows in 30 minutes that would’ve taken days with traditional Actions.”
✍️ htekdev📅 Feb 17, 2026🔗 dev.to

Getting Started with GitHub Agentic Workflows: A Hands-On Guide

Recommended
Built a demo repository with 11 agentic workflows. Practical guide covering gotchas, troubleshooting, and tips to shorten the learning curve. Covers prerequisites, installation, compilation, and deployment.
✍️ Josh Johanning📅 Mar 3, 2026🔗 josh-ops.com

GitHub Agentic Workflows Explained

Recommended
Deep technical analysis of the 3-layer security model (substrate isolation, compilation-time constraints, safe outputs). Covers agent flexibility across Copilot, Claude, and Codex.
✍️ Marina Rivosecchi📅 Mar 23, 2026🔗 tenki.cloud

LinkedIn: GitHub Introduces Agentic Workflows

Supplementary
”A shift from event-triggered automation to agent-driven automation: workflows that can reason about state and adapt their execution.”
✍️ Ilyas F📅 Feb 22, 2026🔗 LinkedIn

SSW Rule: Agentic Workflows

Supplementary
Enterprise software consultancy’s standard rule for when and how to adopt agentic workflows in projects. Practical enterprise perspective.
✍️ Eddie Kranz (SSW)📅 Feb 2026🔗 ssw.com.au
⚡

Quick Reference

Key links at a glance — bookmark these for daily use.

📘
Official Documentation
github.github.io/gh-aw
🗂️
GitHub Repository
github.com/github/gh-aw
📋
Changelog
github.blog/changelog
🚀
Quick Start Guide
gh-aw/setup/quick-start
📝
Markdown Reference
gh-aw/reference/markdown
🔒
Security Architecture
gh-aw/security
💡
Frontmatter & Schema
gh-aw/introduction/overview
🤖
llms.txt (Agent Reference)
gh-aw/blob/main/llms.txt
🔗
Shortlink
gh.io/gh-aw
🛡️
Agent Workflow Firewall
github.com/github/gh-aw-firewall
📦
Sample Workflow Pack
githubnext/agentics
💬
Community Discussions
gh-aw/discussions
🏭

Peli’s Agent Factory Blog Series

19 parts

The definitive blog series by Don Syme and Peli de Halleux (GitHub Next / Microsoft Research). Over 100 workflows built and operated in the gh-aw repository itself. “AI agents are most powerful when they’re specialized, well-coordinated, and designed for their specific task.”

0
Welcome to Peli’s Agent Factory
Introduction — Setting the stage for specialized agent workflows
1
Continuous Documentation Workflows
Part 1 — Automated docs generation, sync, and maintenance
2
Continuous Improvement Workflows
Part 2 — Incremental code and process improvements
3
Continuous Style Workflows
Part 3 — Enforcing consistent coding style with agents
4
Continuous Refactoring Workflows
Part 4 — Agent-driven refactoring and code modernization
5
Continuous Simplicity Workflows
Part 5 — Reducing complexity and keeping code simple
6
Simple Triage Workflow
Part 6 — Automated issue classification and labeling
7
Issue & PR Management Workflows
Part 7 — End-to-end issue and pull request lifecycle management
8
Fault Investigation Workflows
Part 8 — Automated CI failure analysis and root cause investigation
9
Testing Workflows
Part 9 — Automated test generation and coverage improvement
10
Security Workflows
Part 10 — Security scanning, vulnerability detection, and remediation
11
Analytics Workflows
Part 11 — Repository analytics and reporting agents
12
Multi-Phase Improvers (Part 1)
Part 12 — Complex multi-step improvement workflows
13
Multi-Phase Improvers (Part 2)
Part 13 — Advanced patterns for chained improvements
14
Organization & Cross-Repo Workflows (Part 1)
Part 14 — Managing workflows across multiple repositories
15
Organization & Cross-Repo Workflows (Part 2)
Part 15 — Enterprise-scale workflow coordination
16
Advanced Analytics & ML (Part 1)
Part 16 — Machine learning-powered repository insights
17
Advanced Analytics & ML (Part 2)
Part 17 — Predictive analytics and trend detection
18
Project Coordination Workflows
Part 18 (Final) — Orchestrating agents across an entire project
👥

Key People

The creators and contributors behind GitHub Agentic Workflows.

Peli de Halleux (pelikhan)

Creator of gh-aw. GitHub Next / Microsoft Research. Built over 100 specialized agentic workflows documented in the Agent Factory blog series.
🏢 GitHub Next🔬 Microsoft Research

Don Syme (dsyme)

Key contributor and creator of F#. Driving force behind the Continuous AI vision. Featured in Scott Hanselman interview on agentic workflows philosophy.
🏢 Microsoft💎 Creator of F#

Eddie Aftandilian (eaftan)

Engineering Lead for GitHub Agentic Workflows. Overseeing the technical implementation and developer experience.
🏢 GitHub⚙️ Engineering Lead

Idan Gazit

Head of GitHub Next. Quoted extensively in blog posts: “Custom agents for offline tasks, that’s what Continuous AI is.”
🏢 GitHub Next🎯 Head of Research