Sign InOpen Brain
AI EngineerVideoSource Linked

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

GitHub Next’s prototypes move coding agents from personal chat into governed background workflows and shared cloud workspaces that can act on team conversation.

AI Engineer · Aug 8, 2026
Open Source Open MarkdownOpen JSON
Source Summary

Agentic Workflows turns plain-English tasks into Markdown playbooks that compile to GitHub Actions, with permissions in front matter. Ace gives each session a repository branch in a **cloud micro-VM** and can act on shared conversation history.

Practical Implication

Use **Markdown as workflow source** for recurring maintenance such as dependency upgrades, then constrain what agents may read, write, and access. For collaborative work, preserve team discussion as context so people do not have to restate decisions to an agent.

Agent-Ready Context
Agentic Workflows turns plain-English tasks into Markdown playbooks that compile to GitHub Actions, with permissions in front matter. Ace gives each session a repository branch in a **cloud micro-VM** and can act on shared conversation history.

Use **Markdown as workflow source** for recurring maintenance such as dependency upgrades, then constrain what agents may read, write, and access. For collaborative work, preserve team discussion as context so people do not have to restate decisions to an agent.

Agentic Workflows was available in the talk, while **Ace remained a technical preview target**. The demos show a direction for automation and multiplayer development, not evidence about reliability at production scale.
Connected Context · Feed7 Judgment

This combines two previously separate tool directions: declarative, permission-scoped automation for repeatable repository work and conversation-aware cloud sessions for collaborative development. It makes Markdown a reviewable workflow boundary and shared discussion a context source, while narrowing confidence in the multiplayer vision because Ace was still a preview and no production-scale reliability evidence is supplied.

stablyai/orcaOrca already combines isolated workspaces with shared monitoring and remote steering; Ace extends that direction by assigning cloud branches and consuming team conversation as session context.Introducing the new v0 APIThe v0 API reinforces the cloud-runtime pattern with stateful sandboxed execution, but is specialized for app generation rather than the target’s repository automation and collaborative branch workflow.Build from anywhere with Cursor for iOSCursor’s mobile cloud-agent control confirms demand for work that can start away from a local machine, while the target adds repository-native playbooks and explicit permissions as automation controls.Vercel Plugin now available in VS Code and GitHub Copilot CLIVercel’s plugin reduces repeated platform-context setup; the target applies the same context-preservation principle more broadly by carrying shared team discussion into an agent session.
Context Map
toolscoding#coding-agents#cloud-agents#dev-ux
Uncertainty
Agentic Workflows was available in the talk, while **Ace remained a technical preview target**. The demos show a direction for automation and multiplayer development, not evidence about reliability at production scale.