How to Install Hermes Agent on Windows Using WSL – A Step‑by‑Step Guide

1. Install WSL

WSL provides a Linux environment on Windows. Install it with wsl --install. After installation, create a Linux user (e.g., user with password 123456) and close the terminal. Re‑enter the Linux shell later by running wsl.

2. Install Hermes Agent

Method 1 – Clone from GitHub (not recommended)

If GitHub is reachable, run the installer script directly:

curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash

Method 2 – Clone from GitCode (recommended for China)

Clone the repository from GitCode and then execute the same installer script. The installer skips the clone step if the source already exists.

git clone https://gitcode.com/GitHub_Trending/he/hermes-agent.git ~/.hermes/hermes-agent
curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash

The installer automatically pulls Python 3.11, Node.js, Git, uv, creates a virtual environment, clones the repository, and builds Hermes. The process may take several minutes.

3. Run the Configuration Wizard

The wizard prompts you to select an LLM provider (e.g., OpenRouter, OpenAI, Anthropic, DeepSeek) and to enter the corresponding API key. If the wizard fails, edit the configuration file manually. Example .env content:

cat > ~/.hermes/.env << EOF
ANTHROPIC_API_KEY=your_actual_key
HERMES_MODEL_PROVIDER=deepseek   # provider name
HERMES_MODEL_NAME=deepseek-chat   # model name
HERMES_API_BASE_URL=https://api.deepseek.com/v1
EOF

Provider‑specific variable names (e.g., DEEPSEEK_API_KEY) are noted in the wizard.

4. Verify the Installation

Run a health check: hermes doctor The command reports warnings for optional components (e.g., missing tool APIs). These warnings can be ignored until the corresponding features are needed.

5. Start Hermes Agent

Launch the agent with: hermes The agent starts and you can begin chatting with the configured LLM.