# Draught - A provider-agnostic coding-agent CLI and runtime

**URL:** <https://forum.elixirforum.com/t/draught-a-provider-agnostic-coding-agent-cli-and-runtime/76621>\
**Category:** Announcing\
**Tags:** library, cli, llm\
**Created:** [September 8, 2026, 7:35pm UTC](https://forum.elixirforum.com/t/draught-a-provider-agnostic-coding-agent-cli-and-runtime/76621 "2026-09-08T19:35:58Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![mudasobwa](https://forum.elixirforum.com/user_avatar/forum.elixirforum.com/mudasobwa/32/41119_2.png) [@mudasobwa](https://forum.elixirforum.com/u/mudasobwa)\
**Post date:** [September 9, 2026, 6:57am UTC](https://forum.elixirforum.com/t/draught-a-provider-agnostic-coding-agent-cli-and-runtime/76621/4 "2026-09-09T06:57:01Z")

</div>

> [@shahryarjb](#):
>
> why not use OpenCode’s own harness directly on the server?

There is a relatively long [Motivation](https://github.com/ccarvalho-eng/draught#motivation) chapter directly in README.

I can only speak for myself, as a builder of another harness, I find harnesses provided by model providers to be a second-class citizens. Building a harness that would decrease token usage and drastically increase the quality of the outcome does contradict goals of the model providers.

E. g. my harness uses AST for RAG, 100+ MCP tools and whatnot to minimize the necessity of back-and-forth frictions calling the remote. This one aims to do the same, AFAICT.

---

_[View the full topic](https://forum.elixirforum.com/t/draught-a-provider-agnostic-coding-agent-cli-and-runtime/76621)._
