7 Hermes uses you must try (that actually work)

7 Hermes uses you must try (that actually work)

Introduction to Hermes and Claude Code

Overview of Hermes' Capabilities

  • The speaker instructs Hermes to use the Claude Code plugin in VS Code to replicate a competitor's business model, showcasing its automation capabilities.
  • Discussion on how entrepreneurs are utilizing Hermes for various tasks, emphasizing its ability to control applications based on user commands.

Use Cases and Excitement Around Automation

  • The speaker expresses excitement about using Hermes for repetitive tasks at night, highlighting the potential of automation in enhancing productivity.
  • A resolver tool is mentioned that observes daily work patterns and suggests ongoing tasks, indicating a blend of observation and action in workflow management.

Competitor Analysis with Hermes

Building Reports Using Automation

  • An example is given where a report on a competitor is generated automatically by Hermes, providing insights into their technology stack and pricing.
  • The process of converting findings into markdown files for further analysis or replication by Claude Code is discussed as an efficient method for competitive research.

Simplifying Research Tasks

  • A simple prompt demonstrates how easy it is to conduct thorough research using Hermes, which can be set up as a routine task for regular updates.
  • The importance of making these automated tasks repeatable (like cron jobs) is emphasized; one-off tasks often lead to forgotten insights without follow-up actions.

Daily Workflow Management with AI

Organizing Daily Tasks

  • A new morning routine involving triaging daily tasks through a kanban board is introduced, showing how users can delegate responsibilities to their AI agents.
  • Criticism arises regarding the effectiveness of the kanban board interface used with Hermes; some find it lacking context and usability compared to other tools.

Enhancing Productivity Through Prioritization

  • The speaker shares frustrations about managing multiple threads across different platforms; they advocate for better prioritization tools like resolvers that consolidate open assignments.

Combining Tools for Efficiency

Integrating Multiple Platforms

  • Discussion on combining outputs from various tools (Codeex, Cloud Code, etc.) into actionable items highlights the need for seamless integration in workflows.

Personal Experience with Task Management Tools

  • Insights from an AI mastermind event reveal personal strategies for tracking inefficiencies and improving productivity through self-observation techniques.

Communication Strategies with AI Agents

Offloading Tasks Effectively

  • Suggestions include using Telegram or other messaging apps to communicate simple requests directly to Hermes instead of relying solely on complex interfaces.

Creating Automated Logs

  • One participant discusses setting up an automated diary system that logs daily activities via prompts given to their AI agent. This serves as both a memory aid and organizational tool.

Memory Enhancement Techniques

Utilizing External Tools

  • Recommendations are made regarding integrating QMD memory tools with Obsidian for improved data retrieval and organization within workflows.

Continuous Improvement Systems

  • Emphasis on creating closed-loop systems where decisions are logged daily ensures continuous improvement over time through consistent reflection on past actions.

Project Management Integration

Leveraging APIs

  • Discussion around using Linear as a project management tool integrated with other platforms like Cloud Code emphasizes efficiency in tracking progress across projects.

Data Transfer Between Applications

  • Participants share methods of transferring data between applications seamlessly so that insights gained from one platform can trigger actions in another without manual input.

Preferences Between Communication Platforms

  • Debate over Slack versus Discord reveals preferences based on team size; Slack may be more suitable for larger teams while Discord could serve solo users better due to its simplicity.

Video Editing Innovations Using AI

  • Experimentation with video editing software shows potential future capabilities where AI could automate complex editing processes effectively without requiring extensive user input.

AI Agents and Their Applications in Daily Tasks

Overview of Hermes AI Agent

  • The speaker discusses the various tasks performed by their Hermes AI agent, including daily news briefings and YouTube comment monitoring.
  • The Hermes agent is involved in community engagement, business summaries, server checks, research reports, and follow-up reminders.

Daily Briefing Insights

  • The speaker finds the daily briefing useful for identifying relevant news related to AI business and marketing but admits they no longer read it due to information overload.
  • They mention having an "open claw" that runs similar cron jobs but feels overwhelmed by the amount of information generated.

Comment Response Automation

  • A discussion arises about using AI agents to respond to comments on social media platforms like Instagram, which can enhance email list growth.
  • One participant suggests creating downloadable resources quickly through automated responses to drive engagement.

Interviewing with Hermes

  • Charbell demonstrates how his Hermes agent interviews him for better context understanding. This method helps tailor responses more effectively.
  • The importance of asking clarifying questions is emphasized as a way for the agent to gather necessary information before executing tasks.

Skill Development and Workflow Optimization

  • Participants discuss using different skills like "Grill Me" for more thoughtful questioning compared to standard prompts.
  • They share experiences where engaging with the AI leads to generating outlines or content ideas efficiently during casual moments.

Enhancing User Experience with Desktop Applications

Benefits of Desktop Integration

  • Transitioning Hermes to a desktop application allows users greater control over its functionalities and access to skills directly from their main computer.
  • Users can easily manage skills by turning them on or off within the app interface, enhancing usability.

Remote Access Capabilities

  • Discussion includes connecting multiple devices via a remote gateway setup, allowing seamless communication between different machines running Hermes agents.
  • Advanced features enable users to connect various agents across different gateways for enhanced functionality.

Challenges in Task Management with AI

Limitations of Current Tools

  • Concerns are raised about relying solely on file storage without accessible platforms like Notion for organizing outputs from AI tasks.
  • Critique is directed at not providing sufficient background information when instructing the agent, leading to suboptimal results in task execution.

Workflow Efficiency Strategies

  • Emphasis on building workflows that lead directly into actionable steps rather than just generating reports; this includes leveraging APIs effectively.

Skill Creation and Iteration

Importance of Skill Development

  • Participants discuss saving workflows as skills so that repetitive tasks can be executed without re-entering long prompts each time.
  • There’s a debate about whether manual skill updates should be prioritized after completing significant projects versus relying on the AI's ability to self-update based on user interactions.
Video description

Andrew Warner breaks down real Hermes Agent use cases with Eric Siu, from controlling your computer to running research, managing tasks, creating memory systems, and turning repeated work into skills. Presented by Zapier: https://zapier.com/ 👉 Link to resources: https://thenextnewthing.ai/l/best-hermes-use-cases They react to how entrepreneurs are using Hermes for competitor research, daily briefings, YouTube workflows, Slack and Discord communication, Obsidian memory, Linear project tracking, video creation, and remote computer control. Eric also shares what actually works in his own agent setup, what feels like AI theater, and why the most useful workflows are the ones that keep moving work forward instead of creating more reports. Timestamps: 00:00 - Intro 00:27 - Andrew introduces Hermes Agent use cases 01:03 - Eric’s take on loops and agent workflows 01:39 - Competitor research with Hermes 02:51 - Why one-off reports are not enough 03:36 - Kanban boards and AI theater 05:06 - Eric’s Resolver system for open agent work 06:45 - Why Andrew prefers Notion over Hermes Kanban 07:30 - Building a personal wiki with Hermes 08:33 - Obsidian, QMD, GBrain, and agent memory 10:21 - Using Linear for project tracking 11:06 - Passing tasks from Claude to Hermes through Notion 12:27 - Communicating with Hermes through Slack 13:03 - Zapier MCP for email and follow-ups 14:06 - Slack vs Discord for agent communication 16:48 - Using Hermes to create videos with Hyperframes 19:57 - Overlap and AI-assisted clipping workflows 20:33 - Daily briefings, YouTube comments, and crons 22:03 - Using agents for lead magnets and comment workflows 22:57 - Having Hermes ask better clarifying questions 24:36 - Grill Me, interview prompts, and pulling context from yourself 25:48 - Hermes desktop and managing skills 27:27 - Remote gateways and controlling other machines 29:15 - Tailscale and running agents across devices 31:12 - YouTube research workflows 33:00 - Moving beyond reports into actual execution 35:42 - Saving repeatable workflows as skills 36:36 - How Eric decides what to skillify 37:39 - Watch Eric’s deeper agent workflow next 👉 Join us: https://thenextnewthing.ai/