Hermes Agent Framework
Hermes Agent is an open-source autonomous AI agent built by Nous Research and released in February 2026. It is not a coding copilot tethered to an IDE or a chatbot wrapper around a single API. Hermes lives on your server, remembers what it learns, and gets more capable the longer it runs.
Open source under the MIT License. All data stays on your machine. No telemetry, no tracking, no cloud lock-in.
Features
Section titled “Features”Persistent Memory
Section titled “Persistent Memory”Remembers your preferences, projects, and environment across every session. The longer it runs, the better it knows you — no re-explaining context every time.
Automated Skill Creation
Section titled “Automated Skill Creation”When Hermes solves a hard problem, it writes a reusable skill document so it never forgets how. Skills are searchable, shareable, and compatible with the agentskills.io open standard.
Multi-Platform Gateway
Section titled “Multi-Platform Gateway”Connect Telegram, Discord, Slack, WhatsApp, Signal, and CLI — all from a single gateway process. Voice memo transcription, cross-platform continuation. Start a conversation on Telegram, pick it up in your terminal.
Scheduled Automations
Section titled “Scheduled Automations”Built-in cron scheduler with delivery to any platform. Set up daily reports, nightly backups, weekly audits, and morning briefings — all running unattended.
Parallel Sub-Agents
Section titled “Parallel Sub-Agents”Spawn isolated sub-agents for parallel workstreams. Each gets its own conversation and terminal. Collapse multi-step pipelines into zero-context-cost turns via RPC.
Full Browser & Web Control
Section titled “Full Browser & Web Control”Web search, page extraction, full browser automation — navigate, click, type, screenshot. Plus vision analysis, image generation, text-to-speech, and multi-model reasoning.
Capabilities
Section titled “Capabilities”Execution Environments
Section titled “Execution Environments”- Local Terminal — Run commands directly on your machine
- Docker — Isolated container with security hardening
- SSH Remote — Execute on any remote server
- Modal / Singularity — Cloud and HPC execution backends
LLM Providers
Section titled “LLM Providers”- Nous Portal — Native OAuth integration
- OpenRouter — Access 200+ models via API key
- Custom API — Any OpenAI-compatible endpoint
- Local vLLM — Run models fully on-premise
Skills System
Section titled “Skills System”- 40+ built-in skills — MLOps, GitHub, diagramming, note-taking and more
- Auto-created skills — Agent writes new skills as it solves problems
- Community hubs — Browse and install from agentskills.io with one command
- Open standard — Portable SKILL.md format — create, share, reuse anywhere
Security & Privacy
Section titled “Security & Privacy”- No tracking — zero telemetry, zero data collection
- Container hardening — read-only root, dropped capabilities, PID limits
- All memory stored in
~/.hermes/on your machine - MIT License — fully open source, audit every line of code
MLOps & AI Training
Section titled “MLOps & AI Training”Beyond task automation, Hermes Agent is a platform for generating training data, running RL experiments, and exporting trajectories for fine-tuning.
- Batch Processing — Generate thousands of tool-calling trajectories in parallel with automatic checkpointing. Configurable workers, batch sizes, and toolset distributions.
- RL Training — Atropos integration for reinforcement learning on agent behaviors. 11 tool-call parsers for training any model architecture.
- Trajectory Export — Export conversations in ShareGPT format for fine-tuning. Trajectory compression fits training data into token budgets.
Quick Start
Section titled “Quick Start”sudo apt install gnoppix-hermes