# Hermes Agent Features and Practical Usage 🚀 Five minutes from install to your first real conversation. Have the agent read files, run commands, and search the web by handing it an actual task right away. 📌 Title and Feature URL Title: Quickstart URL: 📝 Overview The Quickstart walks you from a fresh install to a working first conversation. You launch with `hermes` or `hermes --tui` and exercise core abilities: inspecting a repo, executing commands, and searching the web. Its guiding principle is to get one clean conversation working before layering on anything else. 🔧 How It Works - Launch via the classic CLI `hermes`, or the modern TUI with modal overlays `hermes --tui` (recommended). - Run the setup wizard with `hermes setup`; pick a model interactively with `hermes model`. - Resume your most recent session with `hermes --continue` (short form `hermes -c`). - Diagnose configuration issues with `hermes doctor`. - Capabilities demonstrated include file reading, terminal command execution (with approval), web search via the Tool Gateway, and multi-turn conversations. 🛠 Practical Usage - Start with specific, verifiable prompts. Example: "Summarize this repo in 5 bullets and tell me what the main entrypoint is." - Confirm file reading with something like "Check my current directory and tell me what looks like the main project file." - Try "Help me set up a clean GitHub PR workflow for this codebase" to see command execution and the approval flow together. - Type `/` mid-conversation for an autocomplete menu of slash commands: `/help` (commands), `/tools` (available tools), `/model` (switch models), `/save` (save the conversation). 🎯 Use Cases - Right after install, double as a smoke test by asking for a repo summary or file inspection. - Feel out the command-approval flow on a real task to calibrate how much you delegate. - Switch models and tools via slash commands as you settle into your workflow. ⚠️ Caveats - Hermes Agent requires a model with at least 64,000 tokens of context; avoid models that are too small. - The guide stresses getting basic chat working first, then adding gateways, skills, or automation. - Command execution involves agent approval, so watch the approval behavior closely on your first run. #HermesAgent# #AIAgents#