Answers from the documentation; checking which model is on duty
Hi - I answer from the OpenSmartRoute documentation: routing, the API, plans and quotas, self-hosting. Ask away, or open a support ticket if you need a person.
Grounded in the docs - follow a source before acting on it.
m3 - Prompt - OpenSmartRoute
Promptv1.0.0
m3
Build, audit or look up a UI against Material 3 (M3 Expressive), mapped onto this shell
Prompt file imported from kagiyanagi/ii-ren (.claude/commands/m3.md). Fill in {{arguments}} before use. Copyright stays with the author.
Follow .agents/skills/m3/SKILL.md in full — it is the single copy of this
procedure, shared with the agy CLI. Treat {{arguments}} as the mode and target it
asks for.
Use it
Copy one of these into your project. Installing also returns the manifest and these snippets.
Follow `.agents/skills/m3/SKILL.md`in full — it is the single copy of this
procedure, shared with the agy CLI. Treat `{{arguments}}`as the mode and target it
asks for.
Manifest
The prompt text and its fill-in variables. Copy it or fetch it by URL from your own code.
kagiyanagi-ii-ren-m3-command.prompt.jsonjson
{
"prompt": "Follow `.agents/skills/m3/SKILL.md` in full — it is the single copy of this\nprocedure, shared with the agy CLI. Treat `{{arguments}}` as the mode and target it\nasks for.",
"variables": [
"arguments"
],
"notes": "Build, audit or look up a UI against Material 3 (M3 Expressive), mapped onto this shell Arguments: build <what> | audit <surface or path> | spec <topic>.",
"metadata": {
"source": {
"provider": "github-claude-commands",
"repository": "https://github.com/kagiyanagi/ii-ren",
"path": ".claude/commands/m3.md",
"ref": "7890bb176d4c7494be3ec9c0b84049a7204d7b14",
"url": "https://github.com/kagiyanagi/ii-ren/blob/7890bb176d4c7494be3ec9c0b84049a7204d7b14/.claude/commands/m3.md",
"key": "kagiyanagi/ii-ren/.claude/commands/m3.md"
}
}
}
Installed into a catalogue, chosen by a router
Install m3 and it becomes one more candidate the router can pick - when it fits.
A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.