# Here’s Google’s Replacement for Vibe Coding Source: https://www.youtube.com/watch?v=PB9sJnZyQ7g Recap page: https://rapidrecap.app/video/PB9sJnZyQ7g Generated: 2025-12-29T17:00:53.011+00:00 --- ## Quick Overview The video details the creation and execution of a complex, context-driven development workflow using the Gemini CLI extension called Conductor, demonstrating how it scaffolds a new Next.js project named Sprint Cycle, defines product requirements, establishes a tech stack, and generates initial tracks for development, despite encountering and manually correcting errors related to database connectivity and missing configuration steps. **Key Points:** - The video showcases the Conductor extension for Gemini CLI, a structured workflow tool for agentic development, contrasting it with standard chat sessions by using Markdown files to persist and track project progress across sessions. - The demonstration involves setting up a new project called 'Sprint Cycle,' a project tracker designed for small to medium-sized software development teams, emphasizing speed and simplicity. - Conductor guides the user through four setup phases: Project Discovery, Product Definition (using Q&A to refine product.md and product-guidelines.md), Configuration (selecting a tech stack, confirming Next.js/TypeScript/Supabase/Drizzle stack), and Track Generation. - Initial track generation for scaffolding failed because the AI incorrectly tried to use 'npm' instead of 'pnpm' (as specified later) and failed to find temporary backup files, requiring manual intervention to delete backups and reconstruct the directory structure. - The AI initially proposed an overly broad single track for building the MVP, which the user corrected, guiding the AI to break it down into smaller, manageable tracks, starting with project scaffolding and core schema. - Despite successful scaffolding and configuration of code style guides and workflow, a subsequent attempt to push the database schema failed due to a missing 'postgres' role, requiring the user to manually check and correct the environment configuration (.env.local file). - The process highlights that while Conductor automates much of the setup, complex projects still require user guidance (like correcting package manager choice and fixing environment variables) to navigate agent errors and ensure correctness. ![Screenshot at 00:02: The terminal screen displays the Conductor workflow in action, showing the Gemini agent executing a 'Shell find' command to locate the conductor directory, indicating the start of the context-aware project setup process.](https://ss.rapidrecap.app/screens/PB9sJnZyQ7g/00-00-02.jpg) **Context:** This video demonstrates the capabilities of 'Conductor,' a specialized Gemini CLI extension designed for structured, context-driven development workflows, which aims to replace simple back-and-forth chat interactions with a persistent, file-based system using Markdown files to define project scope, configuration, and execution tracks. The demonstration focuses on bootstrapping a new web application project, 'Sprint Cycle,' using a modern stack including Next.js, TypeScript, Supabase, and Drizzle ORM, illustrating both the power of the agentic workflow and the necessity of human oversight when the AI encounters configuration issues. ## Detailed Analysis The video presents a deep dive into using the Conductor Gemini CLI extension for developing a project called Sprint Cycle, emphasizing its context-driven, file-based workflow over standard chat interactions. The setup begins with running `gemini --yolo` to initiate the process. Conductor follows four main steps: Project Discovery (analyzing the directory), Product Definition (where the user answers questions to define product.md and product-guidelines.md), Configuration (selecting the Next.js/TypeScript/Supabase/Drizzle stack), and Track Generation. The initial track, 'scaffolding_20251226,' was created successfully after initial errors were resolved. These errors included the agent attempting to use 'npm' instead of 'pnpm' and failing to locate temporary backup files, forcing the user to manually guide the agent to clear old backups and reconstruct the directory structure. Once the groundwork was laid, the agent created the product guide and tech stack files. Subsequent steps involved configuring code style guides and selecting a default workflow, which included 80% code test coverage and task-based commits. When attempting to initialize the database using `pnpm db:push`, the process failed because the environment variable `NEXT_PUBLIC_SUPABASE_ANON_KEY` contained a dummy key, leading to a PostgreSQL error ('role "postgres" does not exist'). The agent correctly identified the need to check the connection string, but the user still had to manually correct the environment file before proceeding. The demonstration concludes by showing that even with a sophisticated system like Conductor, manual validation and correction are essential for complex setups. ### Conductor Workflow Overview - Conductor uses Markdown files to plan and track progress, enabling pause/resume work across machines - Setup begins with `/conductor:setup` which checks existing state via setup_state.json - The process involves Project Discovery, Product Definition, Configuration, and Track Generation. ### Product Definition & Tech Stack - The user defined Sprint Cycle for small-to-medium development teams, emphasizing simplicity and speed - The chosen tech stack was Next.js (App Router), TypeScript, Tailwind CSS, Drizzle ORM, and Supabase (PostgreSQL + Realtime). ### Initial Track Generation Failures - The agent attempted to use 'npm' instead of 'pnpm' and failed to find temporary backup files, requiring manual intervention to clear backups and reconstruct the conductor directory structure. ### Post-Scaffolding Errors - After establishing the initial track, the agent failed to push the database schema due to a 'PostgresError: role "postgres" does not exist' because the environment file contained a dummy Supabase key, requiring the user to manually correct `.env.local`. ### Workflow & Task Management - Conductor utilizes a status indicator in `tracks.md` where `[ ]` is Pending, `[~]` is In Progress, and `[x]` is Completed - The agent generates specific tracks for each major feature, such as 'scaffolding_20251226', which contains `plan.md` and `spec.md` detailing objectives and technical requirements. ![Screenshot at 00:00: The YouTube channel page for 'AI LABS' is displayed, showing recent videos focused on AI tools like Claude and Gemini.](https://ss.rapidrecap.app/screens/PB9sJnZyQ7g/00-00-00.jpg) ![Screenshot at 00:02: The terminal shows the AI agent executing a command within the Conductor workflow, specifically analyzing architecture requirements using a complex prompt structure.](https://ss.rapidrecap.app/screens/PB9sJnZyQ7g/00-00-02.jpg) ![Screenshot at 00:06: A Google Developers article titled 'Conductor: Introducing context-driven development for Gemini CLI' is highlighted, setting the context for the workflow being demonstrated.](https://ss.rapidrecap.app/screens/PB9sJnZyQ7g/00-00-06.jpg) ![Screenshot at 01:11: The user is shown the required command to install the Conductor extension: gemini extensions install https://github.com/gemini-cli-extensions/conductor.](https://ss.rapidrecap.app/screens/PB9sJnZyQ7g/00-01-11.jpg) ![Screenshot at 02:28: The command line interface shows the AI drafting the product guide for 'Sprint Cycle' by asking a series of structured questions about users, goals, and features.](https://ss.rapidrecap.app/screens/PB9sJnZyQ7g/00-02-28.jpg)