@ShenSeanChen: Hermes Agent by just crossed 200K GitHub stars in just a few months. I tested every claim in the @NousResearch project:…

X AI KOLs Timeline Tools
open-source ai-agent memory-system nous-research github hermes-agent

Summary

A tweet thread reports that Hermes Agent, an open-source AI agent framework from Nous Research, surpassed 200K GitHub stars in months. The author tested its memory, skills, cron, and sub-agent features, highlighting that it uses plain text files and a local database for memory, with a growing compound effect.

Hermes Agent by just crossed 200K GitHub stars in just a few months. I tested every claim in the @NousResearch project: memory, skills, cron, sub agents, all from Hermes Desktop App and WhatsApp. The whole system is one loop: message in → agent run → tools fire → loop ends → memory saved → skills grow What surprised me: No embeddings anywhere. Skills and user facts are plain text files. Chat history is one local database. And it still compounds: the longer it runs, the harder it is to leave. The moat is memory, not the model. I caught two gaps too: it never saved a non-complicated skill without being asked, and a cron bug. Details in the video. Watch it, then save the harness diagram below. You Can Build Anything. You Can Learn Anything.
Original Article
View Cached Full Text

Cached at: 07/06/26, 08:07 AM

Hermes Agent by just crossed 200K GitHub stars in just a few months. I tested every claim in the @NousResearch project: memory, skills, cron, sub agents, all from Hermes Desktop App and WhatsApp.

The whole system is one loop: message in → agent run → tools fire → loop ends → memory saved → skills grow

What surprised me: No embeddings anywhere. Skills and user facts are plain text files.

Chat history is one local database. And it still compounds: the longer it runs, the harder it is to leave. The moat is memory, not the model.

I caught two gaps too: it never saved a non-complicated skill without being asked, and a cron bug. Details in the video.

Watch it, then save the harness diagram below.

You Can Build Anything. You Can Learn Anything.

Full 20-min version: https://youtube.com/watch?v=LqG1q5NpOBE…

Chapters: 0:00 What is Hermes Agent 1:17 The full harness diagram 3:00 SOUL.MD: the agent’s system prompt 4:31 Loop engineering + tools 5:23 Live tests: cron, terminal, browser 9:49 Memory system: skills.md + memory.md 13:51 Creating a skill 14:31 Sub agents + Claude Code CLI 16:26 The gaps I found 19:15 Why memory is the moat

Very cool

True

Awesome! Hope you’ll enjoy:))

真的

Makes sense actually. I feel like RAG is for larger scale of memories.

Thanks! I think for smaller memory requirement, text is better coz it’s straightforward. When you have a ton of data, just using text is impossible almost.

Not it’s just on my laptop

Interesting. I think that makes sense if the memory is not large yet

Appreciate it

Similar Articles

@libapi_: 35 days, 4,833 stars. Growth rate far exceeded our expectations. Special thanks to the Hermes official platform for support and recognition. Because of you, we have the opportunity to bring this project to more AI Agent users. This is just the beginning. I'm submitting my homework @NousResearch @…

X AI KOLs Timeline

An AI Agent project called Hermes gained 4,833 GitHub stars in 35 days, with growth exceeding expectations. The team thanks the Hermes official platform and calls for community involvement.