# Realtime multiplayer, automation, and you! — Idan Gazit, GitHub

## Executive summary

The talk introduces two prototypes for future software development: Agentic Workflows and ACE. Agentic Workflows automates complex tasks like dependency upgrades (e.g., Astro 5 to Astro 7) by interpreting plain English instructions into a structured playbook, which is then executed as an action workflow. Crucially, it emphasizes that robust guardrails are defined deterministically in YAML front matter, not merely through prompting. ACE explores real-time multiplayer development in cloud microVMs, treating the shared surface (like Slack) as the primary interface for surfacing non-code facts and collaborative planning.

## Key takeaways

- Automation via Plain English Playbooks: Agentic workflows translate simple natural language instructions (e.g., a message to a junior developer) into comprehensive playbooks that handle tasks like checking for new releases, reviewing changelogs, applying code changes, and creating pull requests.
- Guardrails Must Be Deterministic: Effective security requires defining guardrails (permissions, allowed tools, network destinations, safe outputs) deterministically in front matter (YAML), rather than relying on prompt instructions, which are susceptible to injection.
- Shifting Development Interface: The future of development involves iterating on direction and planning within a shared surface (like Slack/ACE), making the document itself—the 'truth'—a primary artifact, rather than solely relying on code.
- AI Augmentation is Not Typing: A longitudinal study found that hands-on keyboard typing accounts for only about 5% of a developer's time; AI must therefore help scale up the remaining 95% of work (e.g., system design, planning, and collaboration).

## Technical details

- Agentic Workflows Architecture: The workflow is defined by a Markdown document (the source code) which gets recompiled into an executable GitHub Actions YAML artifact. The process includes steps for checking dependencies, reviewing release notes, applying upgrades, and verifying the build.
- Security Guardrails Implementation: Guardrails are specified in a dedicated YAML front matter block to deterministically control permissions (e.g., `read all`), allowed tools, network requests (restricting access only to specific domains like NPM ecosystem), and safe outputs (e.g., allowing only 'pull request single').
- ACE Prototype (Multiplayer Dev): ACE runs sessions in cloud microVMs, ensuring that every session is isolated and not dependent on the local machine. This allows multiple teammates to collaborate on a single repository branch simultaneously.

## Practical implications

- Build engineers can leverage agentic workflows to automate complex maintenance tasks, such as dependency upgrades and build verification, reducing manual intervention.
- The concept of defining guardrails in YAML front matter provides a model for securing CI/CD pipelines against unauthorized actions or prompt injection.
- Adopting shared document editing (Markdown) as the primary source of truth can streamline planning and requirements gathering before coding begins.

## Topics

AI Agents, DevOps Automation, Software Architecture, Collaboration Tools, CI/CD Security, GitHub Next, Astro

Source: https://www.youtube.com/watch?v=iQ5xldZ9StU