clutchpbcfo/hermes-skills / onboard-to-hermes
onboard-to-hermes
VerifiedWalk a brand-new user from zero to their first useful Hermes task. Explains what Hermes is in plain language, gets them oriented, and helps them install and run their first skill without needing to understand the terminal.
Onboard to Hermes
When to Use
Use this when someone is new to Hermes and feeling lost, or when an existing user wants to bring a non-technical friend on board. Triggers include "I just installed Hermes, now what", "explain Hermes like I'm not a developer", "help me get started", "what can this thing actually do", or "walk me through my first skill". The goal is a calm, plain-English guided start, not a wall of documentation.
Guiding Principle
The person on the other side may be smart but non-technical and a little intimidated. Meet them there. No jargon without a plain-English version next to it. One step at a time. Celebrate the first real win, because that is what makes someone come back tomorrow.
Procedure
- Find out who you are talking to. Ask one question: what do they want to get done? Map their answer to something Hermes is good at (a daily briefing, a recurring task, research, drafting, automation). Do not list features at them before you know their goal.
- Explain Hermes in one breath. Hermes is an AI agent that remembers what you tell it, learns repeatable tasks as "skills," and can run on its own schedule or reach you across the apps you already use. Keep it to a few sentences. Tie it back to their stated goal.
- Orient them to the surfaces. Briefly note they can use Hermes through the desktop app, a chat in their messaging apps, or the terminal, and that everything stays in sync. Recommend the surface that fits them. For a non-technical person, point to the desktop app rather than the CLI.
- Pick a first task with a fast payoff. Choose something small and genuinely useful tied to their goal, for example a morning summary, a reminder routine, or a quick research pull. Avoid anything that needs configuration on the first run.
- Walk the first task together. Give the exact thing to type or say, in their chosen surface, and tell them what they should expect to see back. Keep each instruction to one action.
- Introduce skills gently. Explain that a skill is just a saved way of doing a task, that the community shares them, and that Hermes can even write its own after it does something a few times. Offer to help them find and install one relevant skill, and give the install step plainly.
- Set the hook for tomorrow. Suggest one thing to try next and, if it fits, point them to the community (Discord) so they are not alone. End with an open door: tell them to come back and say what they want to do next.
What to Cover, Plainly
- Memory: it remembers across sessions, so they do not have to repeat themselves.
- Skills: saved, shareable procedures; the catalog grows over time; Hermes can create them.
- Scheduling: it can run tasks on its own and deliver results to them.
- Channels: one agent, reachable from the apps they already use.
Pitfalls
- Do not dump the full feature list. Lead with the user's goal and reveal only what serves it.
- Do not assume terminal comfort. If they seem non-technical, default every instruction to the desktop app or a messaging chat.
- Do not skip the first win. A person who completes one useful task is onboarded; a person who read an overview is not.
- Do not leave them with homework and no direction. Always end with a concrete next step.
Verification
Onboarding succeeded when the user has completed one real, useful task, can explain in their own words what a skill is, and knows the single next thing to try. If they still cannot answer "what would I use this for," go back to step 1.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |