7 Hermes uses you must try (that actually work)

Automating Code and Business Analysis [00:00]

• Demonstrated an AI agent, Hermes, controlling VS Code to copy a competitor's business model.
• Highlighted the potential for AI to execute complex tasks like business replication with high accuracy.
• Showcased using AI for in-depth competitor analysis, including technical details, costs, and analytics.
• Emphasized the ease of setting up such analysis with simple prompts.

  All the tech behind it, which is all accurate, but has the payments, how much everything costs.

Enhancing Productivity with Loops and Routines [01:06]

• Discussed the "loops" feature, allowing AI agents to continue tasks until completion.
• Introduced a "resolver" that observes user activity to suggest tasks for looping or continuation.
• Presented a personal routine where tasks are submitted to Hermes for automated execution.
• Expressed skepticism about "AI theater" and the effectiveness of certain Kanban board implementations for AI task management.

  I think this is AI theater. And what I mean by that is it looks like you're doing a lot, but you're actually doing a lot of nothing.

Building Memory and Context for AI Agents [05:09]

• Detailed the creation of a "resolver" to organize and prioritize tasks from various AI tools (Codeex, Cloud Code, Hermes).
• Explained the need for AI agents to have a memory, using tools like Obsidian and QMD to log decisions and improve recall.
• Highlighted the development of personalized AI systems that monitor user activity to identify inefficiencies and suggest skill development.
• Discussed using tools like Gbrain for a curated list of AI skills.

  So, I have there there's a I have QMD which is to T toby's uh memory tool and it and then I have uh there's Gbrain I use Gstack. There's a handful of these uh different repos that I use.

Integrating AI with Project Management and Communication [10:25]

• Showcased integrating AI agents with project management tools like Linear via APIs for task tracking and workflow management.
• Demonstrated using Notion as a Kanban board for tasks, with AI agents creating tasks within it.
• Explored communication methods with AI agents, including Telegram, Slack, and Discord, weighing the pros and cons of each for different use cases (single vs. multiplayer).
• Highlighted Zapier's role in granting AI agents access to email and other services for automated follow-ups and engagement.

  So, for number four, you want to have your Hermes work with a project management product like a Lineion. And I like linear because they have a simple API that you can use and a lot of people use it for building products.

AI-Powered Content Creation and Editing [17:00]

• Demonstrated an AI agent creating a video using "hyperframes" by performing research, installing tools, and analyzing video content.
• Discussed the evolving capabilities of AI in video editing and content generation, predicting future models will make it even easier.
• Shared frustrations with text-heavy video content and advocated for AI to assist in creating more engaging visual content.
• Introduced "Overlap" as a content clipping tool for generating short-form video clips from longer content.

  So I said, "Hey, can you make me a video using hyperframes about what Hermes agent is, how you work, how you remember things?"

Streamlining Workflows with Cron Jobs and Automated Tasks [20:36]

• Showcased various "cron jobs" run by Hermes, including daily news briefings, YouTube comment monitoring, and research reports.
• Discussed the utility of daily briefings for staying updated on AI, business, and marketing news.
• Explored the concept of AI agents replying to comments and the potential for them to analyze and prioritize important feedback.
• Highlighted the strategy of using AI to generate downloadable resources from interviews to grow email lists.

  So, if you guys have noticed on my YouTube videos lately, I've had an AI agent that has access to the transcript and knowledge about me and has been responding to your guys' comments.

Advanced Agent Interaction and Management [23:05]

• Demonstrated how AI agents can "interview" users to gather context and clarify requirements, embedding questionnaires directly within the interface.
• Compared different prompting techniques, such as "Ask User Question" and the "Grill Me" skill, for gathering detailed information.
• Showcased the Hermes desktop app, emphasizing its user-friendly interface for managing and enabling/disabling agent skills.
• Introduced the advanced feature of connecting multiple Hermes agents across different computers using a gateway system, enabling centralized control.

  Use ask user question until you reach clarity. And once you submit that, you're going to see that since this is a desktop app and you're not operating on Telegram, it can do things like embed a questionnaire.

Networking and Remote Agent Control [29:19]

• Explained the advantage of using Tailscale for creating a private network between devices, allowing agents to communicate and control each other remotely.
• Contrasted Tailscale's VPN-like functionality with simple remote screen sharing tools like Jump.
• Shared an experience of remotely resetting a downed AI service using Tailscale and Termius while traveling.
• Emphasized the importance of a network that allows devices to communicate securely from anywhere.

  Yeah, I don't. So screen sharing is one thing but when you have tail scale it's almost like a VPN for like a private network and then all these computers can talk to each other on a private network.

Creating and Managing Reusable AI Skills [35:37]

• Demonstrated saving a complex workflow as a reusable "skill" (e.g., competitor-watch) for agents to execute on new topics.
• Discussed the benefits of creating skills for automating repetitive research and analysis tasks.
• Debated the approach to creating and updating skills, with one perspective favoring manual intervention and another relying on AI to identify and suggest new skills based on observed activity.
• Highlighted the concept of a "skills dojo" for organizing and managing AI skills, and the importance of handing off development to others once core functionality is achieved.

  Save that workflow as a skill called competitor-watch. From now on when I give you a search term, find the top five YouTube videos, save the same info to the file and call out the gaps.