惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Blog — PlanetScale
Blog — PlanetScale
Jina AI
Jina AI
C
Check Point Blog
V
V2EX
H
Help Net Security
Microsoft Azure Blog
Microsoft Azure Blog
P
Proofpoint News Feed
A
About on SuperTechFans
D
DataBreaches.Net
腾讯CDC
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
IT之家
IT之家
WordPress大学
WordPress大学
人人都是产品经理
人人都是产品经理
T
The Blog of Author Tim Ferriss
Recent Announcements
Recent Announcements
Google DeepMind News
Google DeepMind News
云风的 BLOG
云风的 BLOG
MongoDB | Blog
MongoDB | Blog
J
Java Code Geeks
博客园_首页
T
Tailwind CSS Blog
M
MIT News - Artificial intelligence
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻

Google Developers Blog

Why client SDK generation belongs in the open- Google Developers Blog Agent Anomaly Detection, now in Private Preview on the Gemini Enterprise Agent Platform- Google Developers Blog Build zero-trust AI agents that judge intent, not just syntax- Google Developers Blog Autonomous LLM post-training with Tunix on TPUs- Google Developers Blog The Anatomy of Harness Engineering: How to Evaluate, Iterate, and Guard AI Coding Agents- Google Developers Blog Announcing ADK for Kotlin 1.0: Building Production-Ready AI Agents in Kotlin, Android, and Beyond- Google Developers Blog Driving Developer Excellence: Inside the Program Sprints- Google Developers Blog 4 engineering patterns behind the strongest AI Agents Challenge submissions- Google Developers Blog Decoding cosmic signals with deep learning and Keras- Google Developers Blog Enterprise-Grade Precision for Long-Context Multimodal Embedding Inference on Cloud TPU- Google Developers Blog How to Evaluate Live & Voice Agents in ADK- Google Developers Blog Build zero-trust AI agents with Google's Agent Development Kit- Google Developers Blog Introducing Credentio: Open Source C++ Library for C2PA Content Credentials from Google- Google Developers Blog HeyGen x Google Cloud: Bringing Avatar IV to TPUs- Google Developers Blog Why Go is an Ideal Language for AI-Assisted Software Engineering- Google Developers Blog Mastering Edge AI on Raspberry Pi with LiteRT and Gemma- Google Developers Blog Agent Plugins package your skills, tools, and more- Google Developers Blog Scaling AI Agent Infrastructure with the MCP Stateless updates- Google Developers Blog A unified API for AI model routing- Google Developers Blog Scaling real-time AI agents with session-aware load balancing- Google Developers Blog Agent and Model Evaluations in Gemini Enterprise Agent Platform are now GA- Google Developers Blog Enable on-demand expertise with Agent Skills in Genkit Go- Google Developers Blog How to use Google microbenchmarks for evaluating TPU performance- Google Developers Blog Run Ray on TPU, Part 2: Ray AI libraries- Google Developers Blog Scaling Agentic RL: High-Throughput Agentic Training with Tunix- Google Developers Blog Run Ray on TPU, Part 1: The foundations- Google Developers Blog Expanding Choice in Gemini Enterprise Agent Platform: Introducing Grounding with Parallel Web Search- Google Developers Blog Building scalable AI agents with modular prompt transpilation- Google Developers Blog Systems Engineering Playbook: Optimizing Qwen 3.5-397B MoE on Ironwood (TPU7x)- Google Developers Blog Unlocking the Next Era of On-Device AI with Google Tensor and Pixel- Google Developers Blog
Evolving Spec-Driven Development: Conductor Now Supports ...
Mahima Shanware, Sherzat Aitbayev, Jay Kornder · 2026-07-16 · via Google Developers Blog

Mahima Shanware Senior Software Engineer Developer & Experiences

Sherzat Aitbayev Senior Software Engineer Developer & Experiences

Jay Kornder Group Product Manager Developer & Experiences

We introduced Conductor last year to bring spec-driven development directly to your terminal. By shifting project awareness out of ephemeral chat logs and into persistent, version-controlled markdown files, Conductor has helped developers globally plan their architectures before they build.

Today, we are excited to announce the next phase of this journey: Conductor is evolving from a Gemini CLI extension into the Conductor Plugin. Plugins can include skills, rules, MCP servers, and hooks in a single package. Transitioning Conductor into a plugin creates a more fluid, conversational experience without sacrificing the procedural rigor of Spec-Driven Development (SDD). This also means Conductor is compatible with tools such as Antigravity CLI.

Driving SDD via a natural conversation

By transitioning Conductor to a plugin, we are removing the friction of strict command sequences. Conductor will now engage with you conversationally, dynamically generating context, specs, and plans organically as you discuss your feature requirements. You still get the benefits of persistent Markdown artifacts—your spec.md and plan.md aren't going anywhere—but the process of creating and iterating on them is now as simple as chatting with your AI assistant. It intelligently knows when to update your context or check off a completed task in your plan, letting you focus on the architecture while the AI manages the project state in the background.

Conductor for Antigravity CLI and beyond

A benefit of transitioning Conductor to a plugin is that it breaks the workflow out to support multiple tools. Until now, Conductor has been exclusively tied to the Gemini CLI as an extension. By becoming a plugin, Conductor is now a portable, ecosystem-wide capability.

Regardless of which tool you choose, your AI agent will understand the foundational documents about your project's architecture, guidelines, and goals. Your shared configuration and ongoing development tracks will persist seamlessly, meaning an AI-assisted workflow started in one tool can be continued in another without losing a single beat of context.

What this means for your team

  • Backwards compatibility: The new Conductor Plugin works well with old Conductor commands and existing plans and specs explicitly.
  • Natural interaction: Spend less time memorizing commands and more time discussing your implementation strategy.
  • Portability: Use Conductor in the tool that fits your current task, whether that’s Antigravity CLI, Claude, or other tools.
  • Preserved context: Maintain the core promise of spec-driven development. Your repository remains the single source of truth, but how you interact with that truth is now far more intuitive.
  • Improved task completion: The Conductor Plugin achieved a higher success rate across the most complex subset of TerminalBench tasks compared to a user not using SDD.

How to get started

Conductor’s mission has always been to help make AI development safe, predictable, and architecturally sound. By evolving into a conversational plugin, we are making that power more accessible and adaptable to your unique workflow. We can’t wait to see what you build next.

Want to learn more? Test-drive the Conductor Plugin’s features in our hands-on Codelab.

Ready to build? Install the plugin and start using the Conductor Plugin today here, or run the following command to integrate it with Antigravity CLI:

agy plugins install https://github.com/gemini-cli-extensions/conductor