Stop Using Claude Code in Terminal (It’s Holding You Back)
The speaker argues that Claude Code users are being held back by managing multiple terminal tabs instead of focusing on business goals, and introduces a custom command center built on top of their 'Agentic OS' to manage tasks at a higher abstraction level. They critique existing solutions like T-Mux, Anthropic's desktop app, and Vibe Canban for being too developer-focused rather than business-oriented.
Summary
The video addresses a productivity bottleneck that has emerged as Claude Code agents have become more capable. The speaker explains that while Claude Code has improved significantly and can now handle complex autonomous tasks, users are still stuck managing multiple terminal windows, losing context and mental overhead when switching between different agents running simultaneously. The core problem identified is that current interfaces require managing terminals rather than business goals.
The speaker evaluates five main existing solutions and explains their limitations. T-Mux allows multiple terminal panes but still keeps users at the terminal level without big-picture visibility. Anthropic's desktop app offers a cleaner interface but remains limited to one-conversation-at-a-time management and has setup complications for environment variables. Vibe Canban provides a kanban board for coding agents but is designed for developers managing code rather than business owners managing goals. Paperclip offers an operating system for autonomous companies with org charts and roles, but is overcomplicated for simple business tasks. Other tools like Claude Code Board and Mission Control are similarly oriented toward coding sessions rather than business workflows.
The speaker argues that all existing tools approach the problem bottom-up, starting with terminal sessions and trying to add project management layers. Instead, they propose a top-down approach starting with business goals and letting the system determine what sessions to spin up. They introduce their custom-built 'Command Center' that sits on top of their 'Agentic OS' - a business brain containing brand voice, content strategy, and client details that can be deployed in 10 minutes using templates.
The Command Center functions as a kanban board for business goals rather than code management, with iterative workflows between 'your turn' and 'Claude's turn' rather than sequential task management. Users can describe goals in plain business language, set permission levels, and choose task complexity (quick task, campaign, or deep build). The system provides task overviews, conversation history, file outputs, multi-client support, scheduled tasks, skills management, and documentation management all from a single dashboard. The speaker demonstrates how this abstracts away from terminal management to focus on business goal achievement.
Key Insights
- Claude Code agents have become so good at autonomous work that the interface for managing them has become the bottleneck, not the AI capabilities themselves
- Every existing tool approaches the problem bottom-up by starting with terminal sessions and adding project management layers, when what's needed is a top-down approach starting with business goals
- Business users need a level of abstraction similar to hiring a competent employee - give them a goal and deadline, then let them figure out the technical details and provide updates
- Traditional kanban boards don't work for AI agents because human workflows are sequential while AI agent workflows are inherently iterative, requiring constant back-and-forth between user review and AI execution
- None of the existing agent management tools store business context like brand voice, client details, or content strategy - they manage coding sessions in a complete vacuum without understanding the business
Topics
Transcript
[0:00] So, if you've been using claw code for a while, you'll remember the old problem. The output was never good enough. You had to babysit everything, and you'd end up telling it to fix its own mistakes. But now, that's completely changed. Now, the agents are actually good. They can run full tasks. They can handle real workflows and get surprisingly close to what you want. And that's created a new problem. You've got multiple agents running at once, five terminal tabs open, and you're clicking between them trying to remember what each one is doing. So every time you switch, you're losing context. And it's not claw code [0:32] that's slowing you down anymore. It's the way that…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Simon Scrapes
I Rebuilt Hermes in Claude Code (It’s Ridiculously Good)
The creator rebuilds key features of the Hermes agentic AI system inside their own Claude Code setup rather than installing it off-the-shelf, arguing that understanding the architecture provides more long-term leverage. They identify three hidden costs of pre-built systems: inherited assumptions, inability to fix unknown code, and poor scalability. Their custom build addresses Hermes's limitations in multi-client identity management, semantic memory recall, and skill system maintainability.
Agentic AI Systems, Clearly Explained
This video explains agentic AI systems across four levels of complexity: chatbots, AI workflows, agentic workflows, and agentic AI systems. Using content repurposing as a consistent example, the speaker demystifies intimidating terms like harnesses, MCPs, and memory systems, arguing that these systems are fundamentally just files and folders accessible to non-developers.
Stop Downloading Claude Code Skills. Do This Instead.
The video argues that downloaded Claude Code skills are too isolated or bloated to drive real business outcomes. The speaker introduces 'skill systems' — orchestrated chains of small, focused skills that feed into each other end-to-end. A practical example shows a five-skill automation that converts a YouTube video into five ready-to-post short-form clips.
Master 80% of Claude Code, Just Learn These 15 Things
This tutorial covers 15 foundational concepts for using Claude Code effectively, emphasizing that context management is the single most critical skill. The presenter draws from real user behavior across thousands of users to show how plan mode, context tools, slash commands, skills, hooks, MCPs, and memory systems combine to form a powerful agentic workflow.
Master Claude Design in One Video (Complete Build)
The video demonstrates a live build using Claude Design to create real business assets for the Aentic Academy, including a landing page, mobile wireframe, and slide deck. The host emphasizes that building a proper design system upfront is the single most important factor in avoiding generic AI aesthetics. The video compares multiple methods: using ChatGPT's image model, the Skill UI CLI tool, and Claude Design's native interface.