Slice · AI prompt
Authenticated tier-routed LLM access through one Convex action and OpenRouter. Ships a transport-injected ChatFab: React/Next remains default and SvelteKit gets native Svelte 5 UI over the same request/result core and the same aiUsage backend. Unconfigured hosts return an explicit notice instead of a fake reply.
# Install `ai-router` — AI Router — OpenRouter Tier Proxy
> Authenticated tier-routed LLM access through one Convex action and OpenRouter.
📚 Knowledge base : https://resource.rahmanef.com/llms.txt
📦 Slice detail : https://resource.rahmanef.com/slices/ai-router
🧠 JSON catalog : https://resource.rahmanef.com/api/knowledge?slice=ai-router
🔗 Source : https://github.com/rahmanef63/resource-site/tree/main/frontend/slices/ai-router
## 1. Install
```bash
npx rr add ai-router
# alias: npx rr add ai-router
```
The CLI copies `frontend/slices/ai-router/` into your project + augments `.env.example` + installs npm deps automatically. Run it from your project root.
## 2. What it ships
- npm: `convex`, `ai`, `@openrouter/ai-sdk-provider`, `lucide-react`
- shadcn primitives: `button`, `card`, `input`
- env vars:
- `OPENROUTER_API_KEY` (convex, required)
- convex feature paths: `convex/features/ai`
## 3. Wire it up
Run `npx rr add ai-router` or add `--framework sveltekit`. Bind ChatFab.route to api.features.ai.action.callModel. The action requires auth, keeps the provider key server-only, returns a notice when unavailable, and logs successful usage to aiUsage.
## Rules of engagement
- shadcn-only UI primitives. No raw `<button>` / `<dialog>` / native date or file inputs.
- 200-line hard cap per source file (extract neighbours when over).
- All Convex queries hit an index (`.withIndex(...)`); never bare `.collect()`.
- Public mutations/queries declare `args:` validators + authz.
- Full ruleset: https://resource.rahmanef.com/best-practice
The agent will fetch /llms.txt for the full ruleset and use /api/knowledge for the JSON catalog.