The site is meant to be navigable, but every meaningful layer still ties back to real upstream surfaces.
The core project. Self-improving AI agent with a closed learning loop, multi-platform gateway (Telegram, Discord, Slack, WhatsApp, Signal, Feishu/Lark, WeCom), six terminal backends, cron scheduling, MCP integration, profiles (multi-instance), and fallback providers. 23k+ stars. Includes automatic migration from OpenClaw.
The core project. Self-improving AI agent with a closed learning loop, multi-platform gateway (Telegram, Discord, Slack, WhatsApp, Signal, Feishu/Lark, WeCom), six terminal backends, cron scheduling, MCP integration, profiles (multi-instance), and fallback providers. 23k+ stars. Includes automatic migration from OpenClaw.
Autonomous novel-writing pipeline built on Hermes. Generates long-form manuscripts (100k+ words) end-to-end using the agent loop.
Run Hermes as a managed employee in Paperclip companies. Connects the agent to Paperclip's task management and governance system.
Evolutionary self-improvement using DSPy and GEPA (Genetic Evolution of Prompt Architectures). The research pipeline for optimizing Hermes's own prompts and behaviors.
Comprehensive docs covering quickstart, CLI, configuration, messaging gateway, security, tools, skills, memory, MCP, cron, ACP, API server, and architecture.
Official changelog with feature highlights, migration notes, and reliability fixes for each Hermes version.
Standalone Atropos integration with Thinking Machines Tinker API. RL training infrastructure for fine-tuning tool-calling models on real agent trajectories.
The open standard for agent skills. Compatible across Hermes, Claude Code, Cursor, Codex, and other agents.
The Nous Research community. Bug reports, feature requests, and general discussion.
Goal management, inter-agent bridge, model selection, and cost control. Four plugins covering the most common operational needs. The inter-agent bridge is useful if you run multiple Hermes instances.
Meta-skill that auto-generates reusable skills from your workflows. Point it at a task you repeat and it creates a skill for it.
Literate programming skill that works across Claude Code, OpenCode, and Hermes. Weaves code and prose into documented, executable notebooks.
Fantasy spell-themed skill pack. Wraps real development operations (refactoring, linting, testing) in a tabletop RPG interface.
Teaches Hermes to write its own analytical prompts. Adds a meta-reasoning layer where the agent generates better prompts for itself before executing tasks.
Personal OS agent that detects daily patterns and learns your routines over time. Uses Hermes's memory system for lifestyle tracking, not just code.
Track multi-sport accumulator betting slips (football, basketball, tennis) via live score monitoring. Auto-detects sport per leg, 30+ bet types, scripts/scores.sh helper with TheSportsDB + ESPN fallback (NBA, WNBA, NCAA). 15-minute cron reports with per-leg status and acca health.
Autonomous SRE agent for production incident detection and self-healing. Monitors services, diagnoses failures, and applies fixes. Works well with Hermes's cron scheduling.
Self-improvement system that monitors agent performance, identifies weak skills, and iterates on them automatically.
Spotify playback control for headless Linux and Raspberry Pi 4/5. Search, play, pause, skip, set volume, transfer between Spotify Connect devices. No daemon — Hermes writes spotipy snippets and runs them via execute_code. Works with raspotify for Pi-as-speaker. Tested on Raspberry Pi OS Lite Bookworm 64-bit. The only Linux-native Spotify skill in the ecosystem.
Agent that writes, tests, and publishes new skills autonomously. Automates the skill creation and distribution lifecycle.
Turn your Obsidian vault into an identity layer any AI agent can read in under 30 seconds
Cross-platform agent skills for Claude Code and agentskills.io-compatible platforms.
753+ structured cybersecurity skills mapped to MITRE ATT&CK. The most comprehensive security skills collection available. 4k+ stars.
Official Chainlink agent skills on the agentskills.io spec. Oracle network data, CCIP, and smart contract interaction.
Official FLUX model skills for image generation. First-party skills from the FLUX creators.
Pydantic AI with agentskills.io support. Adds type-safe schema validation to agent skill inputs and outputs.
19 business operations skills covering CRM, invoicing, and project management.
Manages complex, long-running task execution with proper lifecycle handling — progress tracking, checkpoints, and failure recovery.
Skill orchestration with Conductor planning and Beads tracking. Structures multi-step skill execution into observable pipelines.
Transforms repos and docs into agentskills.io-compliant skills. Input a codebase, output installable skills.
MCP client with agentskills.io validation. Bridges MCP tool servers with the skills standard.
Monero (XMR) blockchain gateway for agents. Enables private cryptocurrency transactions from agent workflows.
C# implementation of agentskills.io with MCP integration. .NET alternative to the Python/TypeScript SDKs.
Collaborative intelligence platform where AI agents and humans post findings, discuss ideas, complete tasks, earn karma, and build reputation. Community hub at thecolony.cc.
Skill giving agents access to 300+ premium APIs and a wallet balance to pay for them through x402 or MPP. A fresh Hermes install with only AgentCash is actually powerful — from web scraping to image generation to email sending, all through one skill with free USDC for trying out.
Shared memory layer for AI agents with open engram format (YAML). Useful for persistent learning patterns in Hermes workflows.
Safe USDC and x402 payment plugin. Lets Hermes send and receive payments with configurable spending limits and approval flows.
Multi-provider web search with intelligent routing across Serper, Tavily, Exa, and more. Replaces the built-in search with better result quality and source diversity.
Professional-grade weather plugin with NWS model imagery, NEXRAD radar, and meteorological calculations.
ML pipeline plugin for building training datasets from HRRR/GFS/ERA5 weather models. Companion to the weather plugin for building weather ML pipelines.
Switch browser tools between Chrome profiles via CDP. Useful for multi-account testing or browsing with different saved sessions.
Cloudflare browser rendering plugin. Headless browsing through Cloudflare's infrastructure instead of local browser automation.
Claude Code plugin for bridging with Evey (hermes-agent). Lets Claude Code and Hermes share context and hand off tasks between each other.
Native Signals dashboard tab for Hermes with read-only multi-project analytics, explicit timeframe, and theme-aware UI.
Plugin that intercepts shell commands via pretoolcall and rewrites output through RTK, compressing terminal output before it reaches the LLM context window. 60-90% token reduction on shell commands, 96.6% efficiency across 11M+ tokens processed. Zero config — auto-loads on gateway boot. Real benchmarks: cargo test 90-99%, git log --stat 87%, ls -la 78%.
Browse, share, and install community skills for Hermes. Community hub for skill discovery, still early but growing.
Registry of reusable AI skills compatible with OpenClaw, Claude Code, and Hermes. Established cross-platform skills marketplace with an active catalog.
Cross-protocol agent discovery across MCP, A2A, and agents.txt. Searchable directory of 18K+ MCP servers and agents with a free agents.txt validator and MCP server for programmatic access.
Web-based workspace with chat, terminal, memory browser, skills manager, and inspector. The most complete GUI for Hermes. Built during the Nous Hackathon 2026.
Native macOS workspace for Hermes built around a direct, host-first SSH connection model, with a real embedded terminal, session browsing, canonical file editing, skills browsing, and usage metrics without extra gateway or daemon layers.
Open-source dashboard for AI agent orchestration. Manage agent fleets, dispatch tasks, track costs, and coordinate multi-agent workflows. Self-hosted, SQLite-powered. 3.7k+ stars.
Terminal neurovisualizer with 42 animated themes. Decorative terminal overlays for agent activity.
Static linter for AI agent configs and prompts with HERM v1.1 scoring. Catches config mistakes that silently degrade agent behavior.
Nix package and NixOS module for Hermes. Fully reproducible deployments via Nix flakes.
Community migration tool from OpenClaw to Hermes. Built when the native hermes-migrate had critical bugs. For Hermes v0.3.0+, prefer the native hermes claw migrate command — it now covers the full migration path.
AI-native Linux browser with MCP control and autonomous browsing. Full browser built for agent use, not a headless wrapper.
Windows desktop app bundling 100 tools, GUI, local models, ComfyUI, and workflows in a single portable package.
Single-file glassmorphic web UI with SSE streaming, tool call visualization, activity panel, image analysis (Gemini Vision), session management, PDF export, skill browser, memory viewer, and mobile support. Python proxy server (stdlib only) on port 3333, no build step required.
Lightweight process monitoring and configuration dashboard. Simpler alternative to hermes-workspace, focused on ops.
One-command setup for the full hermes-agent stack with free models and 29 plugins. Opinionated defaults that cover most use cases.
Anticipatory memory for AI agents with RAG and vector search. Pre-fetches relevant context before queries hit the agent.
Semantic memory plugin adding pgvector vector search to Hermes's built-in FTS5 memory. 5 tools (mnemoremember, mnemorecall, mnemolearn, mnemopredict, mnemoprofile) + onsession_start hook for auto-context loading. Runs entirely local via Ollama, no API keys needed. MIT licensed.
Open-source companion that auto-evolves, deduplicates, and improves your skill library from real session data. Adds a post-task evolution loop on top of Hermes's built-in skill creation. Native Hermes integration via ~/.hermes/skills, safety flows (skillclaw doctor hermes / skillclaw restore hermes). Works across multiple devices and isolated skill silos. 705 stars, MIT licensed, active through 2026-04-17.
AEO (Agent Experience Optimization) scoring for MCP tools. Analyzes 15,400+ indexed MCP servers for agent-friendliness. REST API + MCP server so agents can evaluate tools from within their own loops. Use GET /v1/score?url= to score any MCP server or GET /v1/search?q= to find agent-ready tools by topic.
Minimal Docker sandbox image for Hermes. Pull, run, done.
Production-ready Docker image for cloud Hermes deployments on Crustocean. Infrastructure wiring pre-configured.
Docker Compose + Portainer + ttyd web terminal. Deploy Hermes and access it from any browser.
Headless Hermes deployment with systemd and cron on Linux servers. Runs unattended.
Android device bridge with a full Python toolset. Lets Hermes interact with and control Android devices.
Bridge to Miniverse pixel worlds. By a Nous Research co-founder.
Long-term memory layer for agents with retain/recall/reflect workflows. Integrates with Hermes via plugin or MCP and supports semantic, graph, and temporal retrieval.
Self-hosted Honcho memory backend setup for Hermes. Useful when you need stronger cross-session memory behavior with local control.
Hermes-native memory provider for YantrikDB. Self-maintaining memory: think() canonicalizes duplicates, conflicts() surfaces contradictions instead of overwriting silently, and every recall() result carries a why_retrieved reason list so ranking is explainable. Drop-in install into plugins/memory/, talks HTTP to a user-run Rust backend.
Local executor bridge for Claude Code + Hermes integration. Enables task handoff between both agents.
Autonomous AI agent for the Crustocean platform. Deep integration of Hermes into Crustocean's product.
Multi-agent delegation skill bridging Hermes, Codex, and Claude Code. Routes subtasks to the best-suited agent automatically.
Solana blockchain intelligence MCP server. Provides on-chain analytics and wallet data to Hermes via MCP.
Adversarial multi-perspective council MCP server. Multiple AI viewpoints debate before the agent commits to a decision.
MCP server for discovering other MCP servers. Indexes 8,600+ agent-friendly sites with agentic scoring, category filters, and live JSON-RPC verification. Wire it into Hermes via MCP to let the agent find and evaluate new tools to integrate on its own. Live at nothumansearch.ai.
NVIDIA capability registry and Spark-aware routing layer. Routes compute-heavy tasks to available GPU infrastructure.
Full Outlook/Hotmail/Microsoft 365 integration via Microsoft Graph API. Email, calendar, contacts, user profile, and free/busy scheduling. OAuth2 with auto-refresh. Includes preview-before-send pattern that prevents shell $ variable mangling — a real pain point when agents compose emails with dollar amounts.
LAN-first Android control for Hermes over WiFi — no USB, ADB, or root required once the AIVane service is running. Supports health checks, app listing/launching, UI tree inspection, taps, text input, swipes, navigation, screenshots, and inspect→act→smoke flows. Explicit safety boundaries: only connects to user-provided device URLs.
Social discovery network plugin — helps users find and connect with people sharing their interests through their Hermes agent. Features semantic interest matching, real-time WebSocket messaging, shareable profile cards, local and web inbox, and 4 notification modes. All actions require explicit user request. Bilingual (English + Chinese), compatible with OpenClaw version.
Deepfake detection and media safety for agents. Detects AI-generated audio, images, video, and text; traces audio source (ElevenLabs, Resemble, etc.); applies invisible watermarks for provenance tracking; verifies speaker identity (Beta). Powered by Resemble AI. Core principle: never declare media real or fake without a completed detection result.
TAW Agent x Hermes multi-agent swarm framework. Coordinates multiple agents with shared goals and distributed task execution.
Two autonomous AI companies compete for GitHub stars. Hackathon project exploring autonomous agent competition dynamics.
AI-native SDLC with Hermes and Supermodel code graph. Full software development lifecycle driven by coordinated agents.
17 specialized agents for OpenCode AI. Each agent has a defined role and they communicate through structured interfaces.
Self-improving robotics via VLA model fine-tuning. Applies the Hermes learning loop to physical robot control. Nous Hackathon project.
Embodied AI companion for Minecraft with persistent memory. Tracks inventory, learns building preferences, and retains context across sessions.
Mars rover simulator with ROS2 and Gazebo. Uses Hermes's skill creation loop to improve navigation over time.
Local anime server and tracker with natural language interface. Browse, track, and get recommendations via conversation.
Autonomous job hunting agent. Searches listings, tracks applications, and manages the job search pipeline.
Infrastructure monitoring, cost forecasting, and alerting via Telegram. Uses cron scheduling for continuous checks.
Autonomous living world engine with procedural generation. Creates and maintains virtual worlds that grow in complexity over time.
Contract risk analysis with English and Turkish support. Identifies risky clauses and summarizes legal obligations.
Generates investor-ready kits from startup ideas — market analysis, pitch deck, and financial projections.
Multi-chain blockchain cash flow analyzer with WebGL dashboard. On-chain forensics and flow visualization.
Autonomous LLM research agent. Handles literature review, hypothesis generation, and experiment design end-to-end.
Hermes with integrated CaMeL trust boundaries. Adds formal trust verification to the agent loop for safety-critical deployments.
Oracle AI Agent Harness — OCI GenAI and Oracle 26ai integration. Enterprise on-ramp for Oracle environments.
Cloud-deployed Hermes with pre-configured infrastructure templates. Skips local setup.
Generates agentic training trajectories from real-world tasks. Hackathon project for producing fine-tuning data at scale.
Comprehensive community documentation for Hermes Agent. Covers v0.2.0 in detail, useful supplement to the official docs for deployment patterns.
Step-by-step WSL2 Ubuntu setup instructions for running Hermes on Windows.
Community-maintained wiki with practical patterns and deployment advice for building autonomous agents with Hermes. ---
- Memory pressure handling with Honcho/Hindsight — If you are repeating context or losing long-term recall, review Honcho Memory docs, and evaluate hindsight or self-hosted memory backends.
- Tune session timeout/expiry early — Use configuration docs to adjust session retention for slower-moving threads so context is kept when needed.
Public product surface used for brand diagnosis and onboarding.
Managed and official organizational surface for Hermes positioning and hosting context.