Guides for using and customizing Hedos — read here, straight from the repo.
Guides
getting-started
Getting started
Installing Hedos, the first-run scan, and what to do next. Start here.
Read →
models
Models
Installing and managing models: the install browser, pasting a Hugging Face or Ollama reference, where downloads land, gated repos and tokens, and deleting a model.
runtimes
Runtimes
How your models run: the engine that fits each model, the managed Python sidecars, and what to expect on first use.
configuring-models
Configuring models
Per-model customization: parameters, the system prompt, context length, renaming, and the honest boundary where a knob applies on one runtime and is refused on another.
cli
CLI
The `hedos` command-line tool: every command and flag, `--json` output, and running headless over SSH.
gateway
Gateway
The local endpoint: enabling the loopback server, client tokens and scopes, the OpenAI- and Ollama-compatible routes, and what it does and doesn't serve.
themes
Themes
The theme system: the two appearance axes, the TOML palette schema, overriding a built-in theme without a rebuild, and adding a new one.
Project
readme
Overview
What Hedos is: the shelf, the runtimes, the gateway, and the CLI — the whole map in one page.
contributing
Contributing
How to build, test, and open a pull request — logic lives in the kernel, the UI stays a thin shell.
conduct
Code of Conduct
The ground rules for participating in the project and its spaces.
security
Security
How to report a vulnerability privately, and what the project promises in return.
changelog
Changelog
What changed in each release, newest first.
license
License
MIT — free to use, read, and fork. No lock-in, no black box.