Hermes Agent is described in its README as 'the self-improving AI agent built by Nous Research' — an agent with a built-in learning loop that creates skills from experience, improves them during use, searches its own past conversations, and builds a user model across sessions. It's MIT-licensed and developed in the open at github.com/NousResearch/hermes-agent.

Hermes Agent is built by Nous Research. The repo (NousResearch/hermes-agent) lists 300+ contributors, but the vast majority of commits come from Nous Research maintainers — the top contributor alone has 2,500+ commits. It's released under the MIT license.

The official installer is a single curl command from the README: `curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash`. It installs uv, Python 3.11, Node.js, ripgrep, and ffmpeg for you. Full docs at hermes-agent.nousresearch.com/docs/getting-started/installation.

v0.14.0 (May 16, 2026) added a native Windows build in early beta, per the GitHub release notes. WSL2 remains the production-recommended path and is what the standard one-liner targets — Hermes also runs on Linux, macOS, and Android via Termux.

Python 3.11, per the README and the pyproject.toml in the repo. The install script automatically installs Python 3.11 via uv, so you don't need to set it up yourself — git is the only prerequisite.

The README lists Nous Portal, OpenRouter (200+ models), NovitaAI, NVIDIA NIM (Nemotron), Xiaomi MiMo, z.ai/GLM, Kimi/Moonshot, MiniMax, Hugging Face, xAI Grok, OpenAI, and custom endpoints as built-in providers. v0.8.0 added native Google AI Studio, v0.9.0 added native xAI (Grok) and Xiaomi MiMo, v0.10.0 wired in the Nous Tool Gateway for managed tools, v0.11.0 brought AWS Bedrock and GPT-5.5 via Codex OAuth, v0.12.0 promoted LM Studio to first class and added GMI Cloud, Azure AI Foundry, MiniMax OAuth, and Tencent Tokenhub, and v0.14.0 added xAI Grok OAuth (SuperGrok) with a 1M-token context plus NovitaAI as an AI-native cloud option. Run `hermes model` after install to pick one.

Hermes Agent supports 22 messaging platforms as of v0.14.0 (May 16, 2026). The baseline set is Telegram, Discord, Slack, WhatsApp, Signal, Email, and CLI; v0.6.0 added Feishu/Lark and WeCom; v0.9.0 added BlueBubbles (iMessage), WeChat (Weixin), and WeCom callback mode; v0.11.0 added QQBot; v0.12.0 added Microsoft Teams and Tencent Yuanbao; v0.13.0 added Google Chat (the 20th platform); and v0.14.0 added LINE and SimpleX Chat. Each platform is configured through the same `hermes setup` flow.

Hermes Agent itself is free and MIT-licensed — and as of v0.14.0 it ships on PyPI, so `pip install hermes-agent` is enough to get started. The costs you pay are to your chosen model provider (Nous Portal, OpenRouter, OpenAI, etc.). Nous Portal offers free MiMo v2 Pro access per the v0.8.0 release notes, and you can also point Hermes at a local model endpoint via LM Studio or any OpenAI-compatible URL.

The install script clones the repo to `~/.hermes/hermes-agent/` by default and symlinks the `hermes` binary to `~/.local/bin/hermes`. You can override the install directory by setting `HERMES_INSTALL_DIR` before running the script.

Run `hermes update` — the CLI has a built-in updater documented in the official install guide. If you prefer manual control, `cd` into the install directory, `git pull`, and re-run `uv pip install -e ".[all]"` inside the venv.

Yes. The README explicitly supports Android via Termux, and the install script auto-detects Termux and installs the `.[termux]` extra. Same one-liner, no rooting needed.

The README describes it as a closed loop where Hermes 'creates skills from experience, improves them during use, nudges itself to persist knowledge, searches its own past conversations, and builds a deepening model of who you are across sessions.' It's the feature the README uses to differentiate Hermes from a stateless chat assistant.

Skills are reusable, composable capabilities. Hermes ships with native support for the open agentskills.io catalog (referenced in the README) and can also generate and save its own skills from working sessions — the self-writing skill system is part of the learning loop.

Yes. v0.6.0 (March 30, 2026) added an MCP server mode so Hermes can expose its tools to other MCP clients. v0.8.0 (April 8, 2026) added MCP OAuth 2.1 for authenticated MCP connections — both highlighted in the official release notes.

v0.6.0 (March 30, 2026) added an official Docker container — it's listed as a highlight in that release's GitHub notes. v0.14.0 (May 16, 2026) added a PyPI package (`pip install hermes-agent`) as a third supported install path alongside the curl script and Docker. Prior to v0.6.0, the curl script plus a local Python environment was the only supported route.

The README lists seven terminal backends for tool execution: local, Docker, SSH, Singularity, Modal, Daytona, and Vercel Sandbox. They let you run Hermes's commands in an isolated environment ranging from a normal shell to an ephemeral serverless container.

v0.7.0 (April 3, 2026) made memory a pluggable provider interface — third-party backends implement a provider ABC and register through the plugin system. Honcho is the reference plugin with profile-scoped host/peer resolution, and it's used as the dialectic user-model backend described in the README. v0.13.0 (May 7, 2026) added 'Session Durability': conversations auto-resume after gateway restarts via checkpoints v2 and atomic session persistence.

The README quickstart is: `source ~/.bashrc` to reload your shell, then run `hermes` to start chatting, `hermes model` to pick a provider, `hermes tools` to configure tools, and `hermes setup` for the full wizard. All of these are listed in the README's post-install section.

The official repository is github.com/NousResearch/hermes-agent — MIT licensed, ~50k stars, ~6.5k forks, and 3,700+ commits on main at the time of writing. Homepage: hermes-agent.nousresearch.com.

The README links three community channels: the Nous Research Discord (discord.gg/NousResearch), GitHub Issues at github.com/NousResearch/hermes-agent/issues for bugs, and GitHub Discussions on the same repo for longer-form questions and skill recipes.

Remove the CLI symlink and the install directory: `rm ~/.local/bin/hermes` then `rm -rf ~/.hermes/hermes-agent`. If you set `HERMES_INSTALL_DIR` during install, remove that path instead. No system-wide packages are touched.

No. This is a fan-made community site. It is not operated by, endorsed by, or affiliated with Nous Research. All content here is sourced from public material — the GitHub repo, release notes, and the official documentation at hermes-agent.nousresearch.com. Hermes Agent and Nous Research are trademarks of their respective owners.

v0.9.0 (April 13, 2026) introduced a Local Web Dashboard — a browser-based management interface for configuration without needing the terminal. v0.11.0 expanded it with i18n (English/Chinese), live theme switching, a plugin system, and mobile-responsive layout. It runs locally next to your Hermes install; the README and the v0.9.0/v0.11.0 release notes describe how to launch it. Beyond the Dashboard, you can still reach Hermes via the chat gateways (Telegram, Discord, Slack, etc., 22 platforms total as of v0.14.0) or via Hermes's MCP server mode for MCP-compatible editors.

Yes — a Linux VPS is the typical deployment. The official requirements are 2 GB RAM minimum (more if you run local models), ~1 GB disk, and Python 3.11 (installed automatically by the curl script). Run the same one-line installer over SSH. For command isolation on a shared host, enable the SSH or Docker sandbox backend from the README so Hermes's shell tools execute inside a container rather than on your VPS filesystem.

The upstream docs are at hermes-agent.nousresearch.com/docs — that is the source of truth for installation, configuration, tools, channels, and the CLI reference. This community site (hermesagents.net) pulls facts from those docs plus the NousResearch/hermes-agent README and the GitHub release notes. If upstream and this site ever disagree, trust upstream.

No. Several unrelated products use the name 'Hermes' — enterprise security agents, email clients, middleware tools. This site only covers Hermes Agent by Nous Research: the open-source self-improving AI agent at github.com/NousResearch/hermes-agent. If you are looking for a Hermes security product, this is not the right site.

Teams launched in v0.12.0 (March 2026) as the first plugin-shipped messaging platform. v0.14.0 (May 2026) wired the full Microsoft Graph stack end-to-end: Graph auth + client foundation, a webhook listener that receives Teams events, a pipeline plugin runtime, and outbound delivery via the Teams adapter. Register a Microsoft Graph app, paste the credentials into `hermes gateway setup`, and the bot can read messages and post back in any Teams channel, DM, or group.

Both were added as first-class platforms in v0.14.0 (May 16, 2026). LINE runs via the official LINE Messaging API and brings Hermes to Japan, Korea, and Taiwan markets where LINE dominates. SimpleX Chat is the privacy-focused decentralized messenger with no user IDs — Hermes can run on it without exposing identifiers. They bring total messaging support to 22 platforms.

Yes — v0.14.0 introduced `hermes proxy`. Run it once and you get a local OpenAI-compatible HTTP endpoint backed by whichever OAuth provider you're signed into — Claude Pro, ChatGPT Pro, or SuperGrok. Codex CLI, Aider, Cline, Continue, and anything else that accepts an OpenAI-compatible endpoint can hit it. One subscription, every tool, no API key required.

Yes — v0.14.0 made `/handoff` actually transfer sessions live. Mid-conversation, hand off to a different model, personality, or profile and Hermes moves the entire session — every message, every tool call, every piece of context — without dropping anything. Useful for switching from a fast model to a deep-reasoning one mid-debugging, or passing a session between profiles for different parts of a task.

Yes, via `/goal`. You specify a target and success criteria; the agent runs in a loop — worker proposes, a separate judge LLM evaluates — iterating until the judge confirms the criteria are met. v0.14.0 added `/subgoal` so you can append new criteria to an active loop mid-flight without restarting (#25449). The pattern is called the Ralph loop internally; it works best for tasks with concrete, gradable success conditions and is a poor fit for fuzzy goals where the judge can't grade consistently.

v0.14.0 (May 16, 2026) added an early-beta native Windows build that installs via a PowerShell one-liner — no WSL needed. It bundles MinGit (~45 MB, no admin) for shell command execution; the CLI, gateway, TUI, and tools all run natively. WSL2 remains the production-recommended path because it's been battle-tested longer; native Windows is functional for casual use but has ongoing platform-specific fixes. The only Hermes feature that currently needs WSL2 specifically is the browser-based dashboard chat pane, which uses a POSIX PTY.

Use the built-in cron scheduler: `hermes cron add`. You specify a name, a schedule (cron syntax like `0 8 * * MON-FRI` or natural language), a natural-language description of the job, and delivery targets. Hermes spawns an isolated agent session at the scheduled time, runs the job, and delivers the output to your messaging gateway. v0.14.0 added `deliver=all` (#21495) to fan output to every connected platform at once — useful for nightly backups, weekly audits, daily reports.

Yes. The `computer_use` tool drives the host's GUI via a cua-driver backend. v0.14.0 made this work with non-Anthropic models (#21967, #24063) — previously it was locked to Claude. Any vision-capable model (GPT-5, Gemini, Grok-vision) can now take screenshots, click, type, and drag. Focus-safe operations prevent the agent from fighting your active window. The driver refreshes automatically on `hermes update`.

No. Hermes supports OAuth login for Claude Pro, ChatGPT Pro, and SuperGrok — sign in with your existing subscription, no API key, no separate billing. v0.14.0 added `hermes proxy` (#25969) which exposes those OAuth providers as a local OpenAI-compatible endpoint, so tools like Aider, Cline, and Codex CLI can use them too. For other providers (Nous Portal, OpenRouter, NovitaAI, NVIDIA NIM, etc.), API keys remain the standard path.

v0.14.0's debloating wave (#24220, #24515) made heavyweight backends lazy-install on first use — the `[all]` extras dropped everything covered by lazy-deps, so a base install is meaningfully smaller. The cold start wave (#22138 and ~10 related PRs) shaved ~19 seconds off `hermes` launch — the `hermes tools` All Platforms screen alone dropped from 14 seconds to under 1.5. Concrete numbers depend on which extras you install, but the post-v0.14.0 experience is fast where it used to be sluggish.

Run `hermes claw migrate`. The migrator imports your SOUL.md (persona), MEMORY.md/USER.md entries, user-created skills (to `~/.hermes/skills/openclaw-imports/`), command allowlist, messaging platform configs, API keys (Telegram, OpenRouter, OpenAI, Anthropic, ElevenLabs), TTS assets, and workspace AGENTS.md. Two presets: `full` (default, includes secrets) and `user-data` (no secrets). Always run with `--dry-run` first to preview. The setup wizard auto-detects `~/.openclaw` and offers to migrate before first config.

Layered defenses. Container isolation is the primary boundary — the agent runs shell commands inside a sandbox (Docker, Daytona, Modal, SSH, Singularity, Vercel, or local — seven backends total). A command-approval workflow flags dangerous operations (rm -rf, sudo, curl|sh) for explicit user approval. v0.14.0 added three specific hardening fixes: a sudo brute-force block (#23736), three closures of known dangerous-command bypasses (#26829), and tool error sanitization (#26823) that strips injection attempts from error strings before they re-enter the model context. The overall security policy was rewritten in v0.14.0 around OS-level isolation as the primary trust boundary (#20317).