I'm Trying to Vibe Code My Entire Job

Quick Overview

The presenter successfully created and demonstrated an automated workflow using Cloud Code (likely an LLM orchestration framework) to pull YouTube performance data, generate a summary report, create a corresponding Notion page, check a calendar for availability, and draft an email to a colleague named Willie, all executed via the command line.

Key Points: The presenter used a custom-built tool integrated with Cloud Code to automate a multi-step workflow involving YouTube analytics, Notion, and Gmail. The workflow successfully pulled YouTube performance data (e.g., 599,612 views, +23,039 subscribers gained in 26 days of 2025) and generated a summary report. The tool then created a new Notion page containing the YouTube performance report, linking it directly in the terminal output. The automation also checked the Google Calendar for availability and drafted an email to a colleague named Willie summarizing the performance and proposing meeting times. The presenter noted that while the setup was initially complex (requiring API tokens for Google/Notion), the resulting command-line execution is fast and intuitive. The presenter mentioned that this automation replaces a manual process that took several hours per week, highlighting the efficiency gains. The presenter indicated that the next step is to ensure all team members can utilize this capability via a slightly customized version of the agent/tool.

Context: The video features a demonstration of an advanced, custom-built automation pipeline, likely utilizing an LLM agent framework referred to as 'Cloud Code,' running on a macOS terminal. The goal is to automate daily operational tasks related to YouTube growth tracking, including data aggregation, report generation in Notion, and communication via Gmail, all triggered by a single command.

Detailed Analysis

The presenter begins by showing the terminal interface where they initiate a custom demonstration they built the previous night, intending to show how they can automate their daily workflow. The tool, identified as 'Cloud Code v1.19,' starts by asking for permissions and then proceeds to execute a complex series of actions. First, it retrieves YouTube performance data using API access (indicated by the commands referencing Google APIs), displaying a summary table showing metrics like Total Subscribers (51,190), Total Views (1,163,674), and YTD performance for January 1–26, 2025. The analysis highlights strong performance, noting that net subscriber growth is +17,991 and average view duration is 7.7 minutes. Next, the agent is instructed to push this report to a new Notion page. The terminal output confirms the creation of the Notion page, providing a link to it. Following this, the agent is tasked with drafting an email to a colleague named Willie, summarizing the performance, linking to the Notion page, and suggesting meeting times for the week. The process also involves checking Willie's calendar via Google Calendar API access. The speaker notes that while the initial setup was complex, involving configuring various API keys and permissions (like for Stripe, YouTube, Google Calendar, Notion, and Gmail), the resulting command-line execution is fast and intuitive, saving significant manual effort previously spent across multiple platforms like spreadsheets and dashboards.