{"slug": "nousresearch-hermes-agent", "title": "NousResearch/hermes-agent", "summary": "Nous Research released Hermes Agent, a self-improving AI agent with a built-in learning loop that creates skills from experience, improves them during use, and builds a persistent user model across sessions. The agent runs on a $5 VPS, GPU cluster, or serverless infrastructure, supports multiple model providers including Nous Portal, OpenRouter, and OpenAI, and offers native Windows, Linux, macOS, and Android/Termux support. It integrates with Telegram, Discord, Slack, WhatsApp, Signal, and CLI, and includes scheduled automations, delegation, and research-ready features.", "body_md": "**The self-improving AI agent built by Nous Research.** It's the only agent with a built-in learning loop — it 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. Run it on a $5 VPS, a GPU cluster, or serverless infrastructure that costs nearly nothing when idle. It's not tied to your laptop — talk to it from Telegram while it works on a cloud VM.\n\nUse any model you want — [Nous Portal](https://portal.nousresearch.com), OpenRouter, OpenAI, your own endpoint, and [many others](https://hermes-agent.nousresearch.com/docs/integrations/providers). Switch with `hermes model`\n\n— no code changes, no lock-in.\n\nA real terminal interface | Full TUI with multiline editing, slash-command autocomplete, conversation history, interrupt-and-redirect, and streaming tool output. |\nLives where you do | Telegram, Discord, Slack, WhatsApp, Signal, and CLI — all from a single gateway process. Voice memo transcription, cross-platform conversation continuity. |\nA closed learning loop | Agent-curated memory with periodic nudges. Autonomous skill creation after complex tasks. Skills self-improve during use. FTS5 session search with LLM summarization for cross-session recall.\n|\n\n**Scheduled automations****Delegates and parallelizes****Runs anywhere, not just your laptop****Research-ready**\n\n```\ncurl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash\n```\n\nHeads up:Native Windows runs Hermes without WSL — CLI, gateway, TUI, and tools all work natively. If you'd rather use WSL2, the Linux/macOS one-liner above works there too. Found a bug? Please[file issues].\n\nRun this in PowerShell:\n\n```\niex (irm https://hermes-agent.nousresearch.com/install.ps1)\n```\n\nThe installer handles everything: uv, Python 3.11, Node.js, ripgrep, ffmpeg, **and a portable Git Bash** (MinGit, unpacked to `%LOCALAPPDATA%\\hermes\\git`\n\n— no admin required, completely isolated from any system Git install). Hermes uses this bundled Git Bash to run shell commands.\n\nIf you already have Git installed, the installer detects it and uses that instead. Otherwise a ~45MB MinGit download is all you need — it won't touch or interfere with any system Git.\n\nAndroid / Termux:The tested manual path is documented in the[Termux guide]. On Termux, Hermes installs a curated`.[termux]`\n\nextra because the full`.[all]`\n\nextra currently pulls Android-incompatible voice dependencies.\n\nWindows:Native Windows is fully supported — the PowerShell one-liner above installs everything. If you'd rather use WSL2, the Linux command works there too. Native Windows install lives under`%LOCALAPPDATA%\\hermes`\n\n; WSL2 installs under`~/.hermes`\n\nas on Linux.\n\nAfter installation:\n\n```\nsource ~/.bashrc    # reload shell (or: source ~/.zshrc)\nhermes              # start chatting!\n```\n\nIf your antivirus (Bitdefender, Windows Defender, etc.) quarantines `uv.exe`\n\nfrom the Hermes `bin`\n\nfolder (`%LOCALAPPDATA%\\hermes\\bin\\uv.exe`\n\n), this is a **false positive**. The file is Astral's `uv`\n\n— the Rust Python package manager Hermes bundles to manage its Python environment. ML-based antivirus engines commonly flag unsigned Rust binaries that download and install packages.\n\n**To verify your copy is authentic:**\n\n```\n# Install GitHub CLI if needed\nwinget install --id GitHub.cli\n\n# Login to GitHub\ngh auth login\n\n# Run verification\n$uv = \"$env:LOCALAPPDATA\\hermes\\bin\\uv.exe\"\n$ver = (& $uv --version).Split(' ')[1]\n[Net.ServicePointManager]::SecurityProtocol = [Net.SecurityProtocolType]::Tls12\n$zip = \"$env:TEMP\\uv.zip\"\nInvoke-WebRequest \"https://github.com/astral-sh/uv/releases/download/$ver/uv-x86_64-pc-windows-msvc.zip\" -OutFile $zip -UseBasicParsing\ngh attestation verify $zip --repo astral-sh/uv\nExpand-Archive $zip \"$env:TEMP\\uv_x\" -Force\n(Get-FileHash \"$env:TEMP\\uv_x\\uv.exe\").Hash -eq (Get-FileHash $uv).Hash\n```\n\nIf attestation says \"Verification succeeded\" and the last line prints `True`\n\n, you're good.\n\n**To whitelist Hermes:**\n\n**Windows Defender:** Run PowerShell as Admin →`Add-MpPreference -ExclusionPath \"$env:LOCALAPPDATA\\hermes\\bin\"`\n\n**Bitdefender:** Add an exception in the Bitdefender console (Protection > Antivirus > Settings > Manage Exceptions)- Whitelist the\n**folder**, not the file hash — Hermes updates`uv`\n\nand the hash changes every version\n\nFor more context, see the upstream Astral reports: [astral-sh/uv#13553](https://github.com/astral-sh/uv/issues/13553), [astral-sh/uv#15011](https://github.com/astral-sh/uv/issues/15011), [astral-sh/uv#10079](https://github.com/astral-sh/uv/issues/10079).\n\n```\nhermes              # Interactive CLI — start a conversation\nhermes model        # Choose your LLM provider and model\nhermes tools        # Configure which tools are enabled\nhermes config set   # Set individual config values\nhermes config get   # Print individual config values\nhermes gateway      # Start the messaging gateway (Telegram, Discord, etc.)\nhermes setup        # Run the full setup wizard (configures everything at once)\nhermes claw migrate # Migrate from OpenClaw (if coming from OpenClaw)\nhermes update       # Update to the latest version\nhermes doctor       # Diagnose any issues\n```\n\nHermes works with whatever provider you want — that's not changing. But if you'd rather not collect five separate API keys for the model, web search, image generation, TTS, and a cloud browser, ** Nous Portal** covers all of them under one subscription:\n\n**300+ models**— pick any of them with`/model <name>`\n\n**Tool Gateway**— web search (Firecrawl), image generation (FAL), text-to-speech (OpenAI), cloud browser (Browser Use), all routed through your sub. No extra accounts.\n\nOne command from a fresh install:\n\n```\nhermes setup --portal\n```\n\nThat logs you in via OAuth, sets Nous as your provider, and turns on the Tool Gateway. Check what's wired up any time with `hermes portal info`\n\n. Full details on the [Tool Gateway docs page](https://hermes-agent.nousresearch.com/docs/user-guide/features/tool-gateway).\n\nYou can still bring your own keys per-tool whenever you want — the gateway is per-backend, not all-or-nothing.\n\nHermes has two entry points: start the terminal UI with `hermes`\n\n, or run the gateway and talk to it from Telegram, Discord, Slack, WhatsApp, Signal, or Email. Once you're in a conversation, many slash commands are shared across both interfaces.\n\n| Action | CLI | Messaging platforms |\n|---|---|---|\n| Start chatting | `hermes` |\nRun `hermes gateway setup` + `hermes gateway start` , then send the bot a message |\n| Start fresh conversation | `/new` or `/reset` |\n`/new` or `/reset` |\n| Change model | `/model [provider:model]` |\n`/model [provider:model]` |\n| Set a personality | `/personality [name]` |\n`/personality [name]` |\n| Retry or undo the last turn | `/retry` , `/undo` |\n`/retry` , `/undo` |\n| Compress context / check usage | `/compress` , `/usage` , `/insights [--days N]` |\n`/compress` , `/usage` , `/insights [days]` |\n| Browse skills | `/skills` or `/<skill-name>` |\n`/<skill-name>` |\n| Interrupt current work | `Ctrl+C` or send a new message |\n`/stop` or send a new message |\n| Platform-specific status | `/platforms` |\n`/status` , `/sethome` |\n\nFor the full command lists, see the [CLI guide](https://hermes-agent.nousresearch.com/docs/user-guide/cli) and the [Messaging Gateway guide](https://hermes-agent.nousresearch.com/docs/user-guide/messaging).\n\nAll documentation lives at ** hermes-agent.nousresearch.com/docs**:\n\n| Section | What's Covered |\n|---|---|\n|\n\n[CLI Usage](https://hermes-agent.nousresearch.com/docs/user-guide/cli)[Configuration](https://hermes-agent.nousresearch.com/docs/user-guide/configuration)[Messaging Gateway](https://hermes-agent.nousresearch.com/docs/user-guide/messaging)[Security](https://hermes-agent.nousresearch.com/docs/user-guide/security)[Tools & Toolsets](https://hermes-agent.nousresearch.com/docs/user-guide/features/tools)[Skills System](https://hermes-agent.nousresearch.com/docs/user-guide/features/skills)[Memory](https://hermes-agent.nousresearch.com/docs/user-guide/features/memory)[MCP Integration](https://hermes-agent.nousresearch.com/docs/user-guide/features/mcp)[Cron Scheduling](https://hermes-agent.nousresearch.com/docs/user-guide/features/cron)[Context Files](https://hermes-agent.nousresearch.com/docs/user-guide/features/context-files)[Architecture](https://hermes-agent.nousresearch.com/docs/developer-guide/architecture)[Contributing](https://hermes-agent.nousresearch.com/docs/developer-guide/contributing)[CLI Reference](https://hermes-agent.nousresearch.com/docs/reference/cli-commands)[Environment Variables](https://hermes-agent.nousresearch.com/docs/reference/environment-variables)If you're coming from OpenClaw, Hermes can automatically import your settings, memories, skills, and API keys.\n\n**During first-time setup:** The setup wizard (`hermes setup`\n\n) automatically detects `~/.openclaw`\n\nand offers to migrate before configuration begins.\n\n**Anytime after install:**\n\n```\nhermes claw migrate              # Interactive migration (full preset)\nhermes claw migrate --dry-run    # Preview what would be migrated\nhermes claw migrate --preset user-data   # Migrate without secrets\nhermes claw migrate --overwrite  # Overwrite existing conflicts\n```\n\nWhat gets imported:\n\n**SOUL.md**— persona file** Memories**— MEMORY.md and USER.md entries** Skills**— user-created skills →`~/.hermes/skills/openclaw-imports/`\n\n**Command allowlist**— approval patterns** Messaging settings**— platform configs, allowed users, working directory** API keys**— allowlisted secrets (Telegram, OpenRouter, OpenAI, Anthropic, ElevenLabs)** TTS assets**— workspace audio files** Workspace instructions**— AGENTS.md (with`--workspace-target`\n\n)\n\nSee `hermes claw migrate --help`\n\nfor all options, or use the `openclaw-migration`\n\nskill for an interactive agent-guided migration with dry-run previews.\n\nWe welcome contributions! See the [Contributing Guide](https://hermes-agent.nousresearch.com/docs/developer-guide/contributing) for development setup, code style, and PR process.\n\nQuick start for contributors — use the standard installer, then work from the\nfull git checkout it creates at `$HERMES_HOME/hermes-agent`\n\n(usually\n`~/.hermes/hermes-agent`\n\n). This matches the layout used by `hermes update`\n\n, the\nmanaged venv, lazy dependencies, gateway, and docs tooling.\n\n```\ncurl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash\ncd \"${HERMES_HOME:-$HOME/.hermes}/hermes-agent\"\nuv pip install -e \".[all,dev]\"\nscripts/run_tests.sh\n```\n\nManual clone fallback (for throwaway clones/CI where you intentionally do not want the managed install layout):\n\nCreate the venv outside the cloned source tree — a venv inside the directory the agent operates from can be wiped by a relative-path command the agent runs against its own checkout, destroying the running runtime mid-session.\n\n```\ncurl -LsSf https://astral.sh/uv/install.sh | sh\nuv venv ~/.hermes/venvs/hermes-dev --python 3.11\nsource ~/.hermes/venvs/hermes-dev/bin/activate\nuv pip install -e \".[all,dev]\"\nscripts/run_tests.sh\n```\n\n- 💬\n[Discord](https://discord.gg/NousResearch) - 📚\n[Skills Hub](https://agentskills.io) - 🐛\n[Issues](https://github.com/NousResearch/hermes-agent/issues) - 🔌\n[computer-use-linux](https://github.com/avifenesh/computer-use-linux)— Linux desktop-control MCP server for Hermes and other MCP hosts, with AT-SPI accessibility trees, Wayland/X11 input, screenshots, and compositor window targeting. - 🔌\n[HermesClaw](https://github.com/AaronWong1999/hermesclaw)— Community WeChat bridge: Run Hermes Agent and OpenClaw on the same WeChat account.\n\nMIT — see [LICENSE](/NousResearch/hermes-agent/blob/main/LICENSE).\n\nBuilt by [Nous Research](https://nousresearch.com).", "url": "https://wpnews.pro/news/nousresearch-hermes-agent", "canonical_source": "https://github.com/NousResearch/hermes-agent", "published_at": "2026-09-02 21:10:21+00:00", "updated_at": "2026-09-02 21:52:48.371886+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-agents", "ai-products", "ai-tools", "ai-infrastructure"], "entities": ["Nous Research", "Hermes Agent", "Nous Portal", "OpenRouter", "OpenAI", "Telegram", "Discord", "Slack"], "alternates": {"html": "https://wpnews.pro/news/nousresearch-hermes-agent", "markdown": "https://wpnews.pro/news/nousresearch-hermes-agent.md", "text": "https://wpnews.pro/news/nousresearch-hermes-agent.txt", "jsonld": "https://wpnews.pro/news/nousresearch-hermes-agent.jsonld"}}