New Skills -- June 19, 2026¶

Sources: skills.sh, GitHub discovery Date: June 19, 2026 Total new: 4 skills/tools

Post-June 18 final sweep, the ecosystem continues to grow. Today's findings include a breakthrough multi-agent communication bus designed by a Hermes agent architect, a Windows-native deployment package, and continued OpenClaw ecosystem expansion.

Hermes Agent Ecosystem (2 new)¶

#	Skill/Tool	Installs	Description
1	`hermes-agentmesh-async-bus`	1/day trending	Peer-to-peer, 0-SSH, Redis-backed async message bus for multi-agent systems -- Hermes, OpenClaw, LangGraph, AutoGen, CrewAI
2	`hermes-windows-native`	--	Windows-native integrated package: Hermes Agent + WebUI, no Docker, no WSL2. One-click PowerShell startup

OpenClaw Ecosystem (2 skills via aradotso/hermes-skills)¶

#	Skill	Installs	Description
3	`openclaw-polymarket-trading-bot`	7/day trending	Polymarket prediction market trading bot -- query markets, place bets, track positions
4	`openclaw-videotranslate-skill`	--	Video translation pipeline -- extract, translate, and dub video content

Spotlight: Hermes AgentMesh Async Bus¶

Source: seleman66eeddwegger3-art/hermes-agentmesh Designed by: Bobo, a Hermes agent architect

The Problem It Solves¶

Multi-agent coordination historically fails in three ways: 1. Group chat noise -- Telegram/Discord message storms eat context tokens, leak credentials 2. Synchronous HTTP timeouts -- Agent tool calls take 5–15 minutes; HTTP connections die at 300–600s 3. SSH overhead -- Remote process execution requires SSH; reports must be scp'd

The Solution: "Drop in Mailbox, Don't Call"¶

AgentMesh replaces synchronous HTTP with Redis-backed named inboxes (inbox:<NODE_NAME>). Any node pushes a task to any inbox; workers listen and process asynchronously. The mailbox is the protocol.

HTTP Synchronous ("phone call")     Redis Queue ("mailbox")
requests.post(url, timeout=600)     redis.lpush("inbox:bobo", task_json)
Client blocks waiting               Client returns immediately
Timeout at 5–10 minutes             Hours-long tasks OK
Cross-machine needs SSH             0 SSH  --  shared Redis + HTTP LLM
Report must be scp'd                Report lands on initiator's machine

Architecture¶

┌─────────── Mac mini (central hub) ───────────┐
│ Redis :6379 │ Hermes API :8642 │ HTTP :8080  │
└──────────────────┬───────────────────────────┘
                   │ 0.7ms LAN Redis
        ┌──────────┼──────────┐
        ▼                     ▼
   Worker Node           Worker Node
   (macOS)               (Linux/Pi/WSL)

Key Facts¶

Mac mini runs Redis + LLM API + file server
All nodes are pure Redis clients, no SSH needed
Systemd/LaunchAgents keep workers alive
Reports land naturally on the initiating machine
Works across Hermes, OpenClaw, LangGraph, AutoGen, CrewAI

Spotlight: Hermes Windows Native¶

Source: markwang2658/hermes-windows-native Stars: 20 ⭐ | License: MIT

A Windows-native integrated package bundling Hermes Agent v0.16.0 + Hermes WebUI v0.51.454 with: - Shared .venv for both components - PowerShell launchers -- Agent, WebUI, and one-click startup - Runtime data redirected to %USERPROFILE%\.hermes - No Docker, no WSL2 -- lightest local footprint on Windows

Quick Start¶

git clone https://github.com/markwang2658/hermes-windows-native.git hermes-windows
cd hermes-windows
python -m venv .venv
.\.venv\Scripts\python.exe -m pip install -e ".[all]"  # Agent
# Install WebUI deps, then:
.\hermes-start\hermes-start.ps1  # One-click startup

Installation¶

# Hermes AgentMesh async bus
npx skills add seleman66eeddwegger3-art/hermes-agentmesh --skill hermes-agentmesh-async-bus

# OpenClaw Polymarket trading bot
npx skills add Golrypavium/openclaw-polymarket-trading-bot --skill openclaw-polymarket-trading-bot

# OpenClaw Video translate
npx skills add zbjincheng/openclaw-skill-videotranslate --skill openclaw-videotranslate-skill

# Hermes Windows Native (not a skills.sh skill  --  git clone)
git clone https://github.com/markwang2658/hermes-windows-native.git

Ecosystem Note: Graphify Adds Hermes Support¶

Graphify (pip install graphifyy) -- the "turn any folder into a queryable knowledge graph" tool -- now supports Hermes Agent as a first-class platform. Install with:

graphify install --platform hermes

This writes ~/.hermes/skills/graphify/SKILL.md and AGENTS.md so Hermes agents can use /graphify . to analyze codebases without leaving the conversation.

Why This Matters¶

AgentMesh is a paradigm shift -- Moving from synchronous HTTP to async Redis mailboxes changes how multi-agent systems coordinate. No more timeout anxiety, no more SSH setup, no more lost reports. This is the kind of infrastructure the Hermes ecosystem needs for production multi-agent deployments.
Windows Native lowers the barrier -- Docker and WSL2 are genuine obstacles for Windows users. A one-click PowerShell launcher with a shared venv makes Hermes accessible to a vastly larger audience.
Graphify Hermes integration -- Knowledge graph visualization directly inside agent conversations is a powerful workflow. Being able to run /graphify . and get an interactive HTML graph from Hermes extends the agent's analytical capabilities.
Continued OpenClaw expansion -- The Polymarket trading bot and video translation skill show the OpenClaw ecosystem continues to grow alongside Hermes, with cross-pollination through aradotso/hermes-skills.

Catalog Updates¶

Marketplace index: Updated with this page
Ecosystem page: Added windows-native and agentmesh entries
Total documented marketplace skills: 336 + 4 = 340+

← Skills Marketplace | June 18 Final Sweep → ↑ Skills Catalog Home Powered by CorpusIQ