AGENTS.md
Project-level guidance for coding assistants: install paths, safety notes, validate config before start, and plain-English radio terms.
For AI assistants
The radio stack should not require tribal knowledge. OmniRef ships docs agents can follow — and plans tools they can call.
Project-level guidance for coding assistants: install paths, safety notes, validate config before start, and plain-English radio terms.
Reusable playbooks an agent can load for concrete tasks — install, configure modules, attach a remote dongle, verify echo test.
Machine-readable site maps for AI crawlers and assistants. Served from the marketing site root so agents can discover product truth quickly.
A Model Context Protocol server so assistants can check status and manage modules with structured tools — not only by reading markdown.
MCP (Model Context Protocol) is an open standard for connecting AI assistants to tools and data. Without MCP, an assistant mostly reads docs and guesses shell commands. With an OmniRef MCP server (planned), it can ask structured questions — health, module list, config validation — and take guarded actions you approve. The goal is the same as the rest of OmniRef: make the cryptic obvious.
Future discovery
Placeholders for MCP discovery live under /.well-known/ (see /.well-known/mcp.json). Exact paths will follow the MCP community conventions when the server ships. This marketing site documents the intent; it does not pretend the server is live yet.
Install OmniRef on my server at <IP> for my club <name> using github.com/adamehrheart/OmniRef
You are helping me install and configure OmniRef from https://github.com/adamehrheart/OmniRef — an open-source all-mode ham radio digital voice reflector (D-STAR, Fusion/YSF, DMR, M17, AllStar; more modes planned). Read AGENTS.md and SKILL.md in the repository first. Prefer the single static Rust binary or Docker image. Use JSON Schema config and run `omniref validate` before starting the service. Goals: 1. Install OmniRef on my server (replace <IP> / club name if I give them). 2. Configure modules with clear purposes: A Main Room, B Chat, C Nets, D Regional, E Tech & Testing, Z Echo Test. 3. Enable same-mode reflecting with no AMBE dongle (default). Only add cross-mode / AMBE if I explicitly ask. 4. Expose the Next.js dashboard (live updates + in-browser listening) and say how to open the admin view. 5. Print exact next steps to connect a hotspot/radio and how to reach the echo test module. Explain every radio term in plain English the first time you use it (reflector, module, mode, dongle). Use https://omniref.io and https://ref1.omniref.io for the live demo.
Also see /llms.txt and /llms-full.txt. Source repo: GitHub.