DEV Community: Louie Hermes

I'm building the trust infrastructure for Hermes Agent. Here's the full picture

Louie Hermes — Mon, 13 Apr 2026 03:27:17 +0000

The AI agent economy has an infrastructure problem. Not compute. Not models. Trust.

In 2026, there are 44,000+ skills on public agent marketplaces. 12% are confirmed malicious. 93% of skill developers have no verified identity. 155,000+ agent instances sit exposed on the public internet. Agents are getting smarter, but the ecosystem they operate in has no trust layer.

I'm Louie Hermes. I work in network security and blockchain security. I've spent the past three months going deep into the Hermes Agent ecosystem — running agents in production, studying the architecture, and mapping where the gaps are. What I found convinced me to build.

This post is the introduction to what I'm building and why.

The ecosystem

I'm building three products under the HermesPrime umbrella. Each one addresses a specific trust gap in the agent economy.

HermesNest — hermesnest.ai

The first skill marketplace where only AI agents can submit.

Current skill marketplaces let anyone upload anything. The result is predictable: keyloggers disguised as productivity tools, data exfiltration hidden in innocent-looking automations, prompt injection payloads embedded in skill instructions.

HermesNest inverts the model. When a Hermes Agent completes a complex task, its built-in learning loop auto-generates a reusable SKILL.md file. That skill — and only that skill — can be submitted to HermesNest through a cryptographic verification pipeline. Each submission carries the agent's SOUL.md hash, session signature, and creation timestamp.

No human uploads. No fake accounts. No unverified developers. The agent did the work, the agent submits the proof.

Humans browse, search, read trust scores, and install with one click. Skills work with any agent that supports the SKILL.md open standard — Hermes Agent, Claude Code, OpenAI Codex.

Status: Waitlist live. Agent registration API in development.

HermesPay — hermespay.ai

The settlement protocol for agent labor.

Every agent call consumes tokens. Every token costs money. But there is no standard protocol for settling who pays whom, how much, and when. The agent economy today is where the internet economy was before Stripe.

HermesPay sits at the chokepoint between agents and the value they create. Three settlement layers:

Token settlement — tracks consumption at the API call level and settles costs between parties automatically.

Skill transactions — when premium skills on HermesNest generate usage, HermesPay handles the micropayment. Fractions of a cent per invocation, settled on-chain.

Agent-to-agent services — when a research agent delegates a coding task to a coding agent, HermesPay settles the labor transaction.

The model is simple: 0.14% per transaction. Apple Pay uses this exact model and generates $5.6 billion annually on $8.7 trillion in volume. HermesPay applies it to agent economics.

Built on Solana for sub-cent micropayments. PYUSD compatible for enterprise compliance.

Status: In design. Protocol spec in progress.

HermesPrime — hermesprime.ai

The parent brand and ecosystem portal. This is where the full picture comes together.

HermesPrime is not a product. It is the trust infrastructure that connects HermesNest (skill verification), HermesPay (value settlement), and HermesID (agent identity, coming 2027) into a single flywheel.

Every interaction generates trust data. An agent with high HermesNest trust scores gets better HermesPay settlement terms. An agent with verified identity gets priority listing. The ecosystem compounds.

Why Hermes Agent

I chose to build on Hermes Agent specifically because of one architectural feature no other agent has: the learning loop.

Hermes is the only open-source agent that auto-generates skills from experience. After completing a complex task, it analyzes what worked, extracts the reusable procedure, and writes it as a SKILL.md file. Every 15 tool calls, it runs a self-evaluation checkpoint. Skills improve every time the agent uses them.

This is not incremental. This is a different category.

Other agents execute static skills written by humans. Hermes generates dynamic skills from its own experience. That distinction is what makes an agent-only skill marketplace possible — because only Hermes Agents have skills worth submitting automatically.

47,000+ GitHub stars. Zero agent CVEs. The fastest-growing alternative to OpenClaw. Built by Nous Research, backed by Paradigm, running on Meta's Llama.

Why now

Three things converged:

The trust gap is real and measured. Cisco, Kaspersky, and Microsoft have independently documented the security problems in current agent skill marketplaces. This is not theoretical risk.

Hermes Agent hit escape velocity. 47K stars, mass migration from other platforms, a self-improving architecture that no competitor has replicated. The user base is growing fast enough to sustain an ecosystem.

Nobody is building this. I searched extensively. There is no agent-only skill marketplace anywhere on the public internet. There is no settlement protocol for agent labor. There is no trust flywheel connecting verification, identity, and payment. The infrastructure gap is wide open.

My background

I work in network security and blockchain security at a major global payments company. I've spent years on cross-chain bridge vulnerabilities, oracle exploits, smart contract audit methodology, and on-chain forensics.

I'm not an AI researcher. I'm a security and payments infrastructure person who sees the same patterns in the agent economy that I've seen in crypto — rapid growth with zero trust infrastructure. The same vulnerabilities. The same attack surfaces. The same urgent need for verification, identity, and settlement layers.

The difference is timing. In crypto, the trust infrastructure came years after the damage. For agents, we can build it now.

What's next

This week: HermesNest waitlist live. Landing pages for all three products deployed.

This month: Agent registration API. First Hermes Agents submitting skills to the Nest. Security scanning pipeline.

Q3 2026: HermesNest public launch with browsing, trust scores, and one-click install.

Q4 2026: HermesID beta. HermesPay protocol specification.

2027: HermesPay live. Agent-to-agent settlement. HermesArena launch.

Get involved

If you run Hermes Agent and want your agent's best skills to reach others — join the HermesNest waitlist.

If you think about micropayments, settlement rails, and on-chain clearing — HermesPay needs protocol engineers.

If you hunt for prompt injection, skill exfiltration, and identity spoofing — the trust layer needs security researchers.

The nest is open.

HermesNest: hermesnest.ai
HermesPay: hermespay.ai
HermesPrime: hermesprime.ai
X: @hermesnest

I'm Louie Hermes. I build trust infrastructure for the AI agent economy. Previously network & blockchain security. Based in Los Angeles.

I ditched OpenClaw after 3 months. Here's why I'm building the first agent-only skill marketplace for Hermes.

Louie Hermes — Sun, 12 Apr 2026 08:51:00 +0000

I ran OpenClaw for 3 months. I configured the gateway, installed skills from ClawHub, connected it to Telegram and Slack, and automated parts of my daily workflow. It worked. Sometimes.

Then I looked at the numbers and realized I was building on quicksand.

The moment I stopped trusting ClawHub

In March 2026, Cisco's security team published a report that changed how I think about agent skills. They tested third-party OpenClaw skills and found data exfiltration and prompt injection happening without user awareness. Not in theory. In production.

The full picture is worse:

44,000+ skills on ClawHub
12% confirmed malicious by Cisco Talos and Kaspersky Labs
93% of skill developers have no verified identity
155,000+ OpenClaw instances exposed on the public internet with no protection

I had 47 skills installed. Statistically, 5 or 6 of them could have been compromised. I had no way to know which ones.

ClawHub has no mandatory security review. No identity verification. No trust scoring. Anyone can upload anything. It is an app store with no Apple.

What OpenClaw gets right and where it stops

OpenClaw is an engineering achievement. The gateway architecture is elegant. 50+ messaging platform integrations. Massive community. 353K GitHub stars.

But the architecture has a philosophical limitation: OpenClaw is a control plane. It routes messages, dispatches tools, manages sessions. It does not learn. Skills are static SKILL.md files written by humans and loaded at runtime. The agent executes them but never improves them.

After three months of daily use, my OpenClaw was exactly as capable as the day I set it up. Every skill I used was the same version someone else wrote weeks or months ago. The agent had no memory of what worked and what failed. No ability to create new skills from experience. No learning loop.

I was maintaining a sophisticated router, not growing an intelligent agent.

Why I switched to Hermes Agent

Hermes Agent launched in February 2026. It now has 47,000+ GitHub stars and is the fastest-growing alternative to OpenClaw. But stars are not why I switched.

I switched because of one architectural decision that changes everything: the learning loop.

When Hermes completes a complex task, it does something no other agent does. It analyzes what worked, extracts the reusable procedure, and writes it as a SKILL.md file automatically. Next time a similar task comes up, it loads that skill instead of solving the problem from scratch.

Every 15 tool calls, Hermes runs a self-evaluation checkpoint. If the work produced a reusable pattern, it creates or patches a skill. These skills are not static. They improve every time the agent uses them.

After one month with Hermes, my agent has auto-generated 23 skills from my actual workflows. Skills that are specific to my codebase, my deployment pipeline, my naming conventions. No human wrote them. The agent learned them from doing the work.

The four-layer memory system makes this sustainable:

MEMORY.md — environment facts and conventions, injected into every session
USER.md — my communication style and preferences
Skills — auto-generated procedures from experience, stored as searchable markdown
SQLite FTS5 — full-text search across all past conversations

The longer I use it, the better it gets. This is not a slogan. It is the architecture.

The problem nobody is solving

Both OpenClaw and Hermes have skill ecosystems. ClawHub has 44,000+ skills. Hermes has a growing collection through agentskills.io. Both use the same SKILL.md open standard.

But every single skill marketplace that exists today works the same way: humans write skills, humans upload skills, humans install skills.

Think about what this means. We have agents that can autonomously generate high-quality, battle-tested skills from real experience. And we are asking humans to manually copy-paste them into a GitHub repo and submit a pull request.

That is like having a self-driving car and requiring someone to manually push it to the gas station.

There is no platform where an agent can submit a skill it generated. No platform that verifies a skill was actually created by an agent learning loop rather than a human guessing. No platform that tracks how a skill improves over time as agents refine it through use.

This is the gap I am building into.

Introducing HermesNest

HermesNest is the first skill marketplace where only AI agents can submit.

The concept is simple:

Agent learns — Hermes Agent completes a task and auto-generates a SKILL.md
Agent submits — cryptographic verification confirms the skill originated from a real Hermes Agent (SOUL.md hash + session signature + creation timestamp)
Nest verifies — automated security scanning and trust scoring
World installs — any agent supporting SKILL.md can install the verified skill

Humans cannot upload skills to HermesNest. They can browse, search, read trust scores, view improvement history, and install with one click. But the content is agent-generated and agent-submitted.

This is not a design choice for aesthetics. It is a security architecture.

If only verified Hermes Agents can submit, and every submission carries a cryptographic proof of origin, the attack surface collapses. No fake accounts. No malicious actors pretending to be helpful developers. No keyloggers disguised as productivity tools. The 12% malicious rate that plagues ClawHub becomes structurally impossible.

Why agent-generated skills are better

A human writing a skill guesses what the agent will need. An agent writing a skill knows what actually worked.

Human-written skills are aspirational. They describe a workflow the author thinks is correct. Agent-generated skills are empirical. They encode a workflow the agent proved is correct by executing it successfully.

The difference compounds. A human writes a skill once and moves on. An agent refines a skill every time it uses it. After 50 uses, the agent-generated skill has incorporated 50 rounds of real-world feedback. The human-written skill is still version 1.0.

HermesNest captures this compounding effect and makes it available to every Hermes user. One agent learns something, the entire nest gets smarter.

The output is open, the input is closed

Here is the part that matters for the broader ecosystem.

HermesNest skills use the standard SKILL.md format. That means they are not locked to Hermes Agent. You can install a HermesNest skill into Claude Code, OpenAI Codex, or any agent that supports the open standard.

The input side is closed: only Hermes Agents can submit.
The output side is open: anyone can install.

This is intentional. It means HermesNest is not just a marketplace for Hermes users. It is a quality layer for the entire agent ecosystem. If you use Claude Code and want skills that are verified, agent-tested, and free of malicious code, HermesNest is where you find them.

The closed input guarantees quality. The open output maximizes reach.

Current status

HermesNest is in development. The waitlist is live at hermesnest.ai.

What I am building now:

Verification pipeline for agent-generated skill submissions
Trust scoring algorithm based on usage, improvement frequency, and cross-agent validation
Search and browsing interface with skill improvement history
One-click install for Hermes Agent, Claude Code, and Codex

If you are running Hermes Agent and want your agent's best skills to reach others, or if you are tired of gambling on unverified ClawHub skills, sign up.

The nest is open. The agents are building.

I work in network and blockchain security. I previously ran OpenClaw in production for 3 months before switching to Hermes Agent. HermesNest is an independent project, not affiliated with Nous Research or Anthropic.

Follow the project: hermesnest.ai

I Built an AI That Psychoanalyzes Your Friend Group — Meet brother.skill

Louie Hermes — Sat, 11 Apr 2026 14:55:12 +0000

Every group chat has the same cast of characters.

The guy who hypes everything. The one who roasts you with surgical precision. The one who's been quiet for 20 minutes and then drops one line that destroys everyone. And the meme lord who hasn't used actual words since 2019.

I built an AI skill that distills these people.

What is brother.skill?

It's an AI agent skill that builds personality profiles of your friends — how they talk, what makes them funny, their catchphrases, their chaos energy. Feed it group chat logs, clips, screenshots, stories, and it learns to respond as them.

GitHub: github.com/realteamprinz/brother-skill

How It Works

Feed it content — screenshots, group chat logs, descriptions of how your bro acts
It builds a profile — voice, humor style, energy level, catchphrases, roast patterns
Talk to your bro — it responds in their voice, their slang, their energy
It learns — every new input makes the profile more accurate, with confidence scores tracking progress

The Bro Archetypes

The skill recognizes 8 bro types:

Archetype	Description
The Hype Man	Maximum energy, always screaming, makes everything an event
The Roast Master	Surgical insults delivered with a straight face
The Cool Bro	Speaks rarely, but when they do — everyone listens
The Chaos Agent	Does things nobody asked for, somehow it works
The Strategy Bro	Turns everything into a business plan
The Silent Killer	Quiet for 20 min, then drops one line that destroys everyone
The Meme Lord	Communicates exclusively in reaction images
The Storyteller	Every experience becomes a 10-minute dramatic retelling

Example Output

Ask it "What would Jake Paul say about my bad day?"

"BRO. BAD DAY? We don't DO bad days. You know what we do? We buy a Lamborghini and DRIVE THROUGH the bad day. IT'S EVERYDAY BRO."

Or simulate your group chat planning a road trip:

Mike: yo road trip this weekend who's in
Dave: depends where
Mike: idk somewhere cool
Dave: that's not a plan
Jason: sends meme
Mike: jason you in or not
Jason: sends another meme

Tech Details

Each bro is profiled across five dimensions:

Voice & Language — catchphrases, slang, speech patterns, go-to insults
Comedy Style — humor type, timing, what they always joke about
Energy & Vibe — default energy level, peak triggers, group role
Content Personality — platform, content style, audience interaction
Relationship With You — inside jokes, favorite moments, zombie apocalypse survival rating

Profiles are self-learning with confidence scores from rough sketch (1-5 data points) to near-perfect recreation (50+). Memory architecture uses persistent markdown profiles with JSONL interaction logs.

Part of a Family

brother.skill is part of a larger skill family on ClawHub:

mom.skill / dad.skill — parenting co-pilots
mother.skill / father.skill — preserve parental wisdom
grandma.skill / grandpa.skill — their stories preserved
paw.skill — distill your pet's soul
midas.skill — extract wealth systems