Name: routerbase-model-routing
Author: yangsonhung

Imported from yangsonhung/awesome-agent-skills (skills/en/routerbase-model-routing/SKILL.md). Install upstream with npx skills add yangsonhung/awesome-agent-skills --skill routerbase-model-routing. Copyright stays with the author.

RouterBase Model Routing

Overview

Use routerbase to plan model selection behind one OpenAI-compatible integration surface. This skill turns workload requirements into a practical model shortlist, fallback chain, cost and latency policy, and validation checklist.

When to Use

Use this skill when the user asks for:

Choosing RouterBase models for chat, coding, reasoning, vision, or multimodal workloads
Designing fallback chains for provider outages or degraded quality
Balancing cost, latency, context length, output quality, and availability
Documenting model routing rules for production applications or agent workflows
Creating a validation plan for model behavior before rollout

Do not use

Do not use this skill for:

Claiming exact pricing, availability, or model performance without checking current source data
Recommending a single model without stating assumptions and fallback behavior
Hiding routing changes from the application owner or production operator

Instructions

Clarify workload type, quality threshold, latency budget, context size, expected traffic, and failure tolerance.
Group candidate models by capability and role: primary, economical fallback, high-quality fallback, and specialized fallback.
State routing assumptions clearly and mark any values that require current catalog or pricing verification.
Recommend small rollout steps: test prompts, golden cases, logging, retry policy, rate limit handling, and rollback criteria.
Keep examples OpenAI-compatible where possible so the application integration stays portable.

Install routerbase-model-routing 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.

{ "ocm": "1", "id": "yangsonhung-awesome-agent-skills-routerbase-model-routing", "kind": "skill", "name": "routerbase-model-routing", "description": "Use when choosing RouterBase models, designing fallback chains, or documenting cost-aware and latency-aware routing policies.", "publisher": "yangsonhung", "version": "1.0.0", "capabilities": { "domains": [ "general" ], "tags": [ "skill-md", "skills-sh" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "Use when choosing RouterBase models, designing fallback chains, or documenting cost-aware and latency-aware routing policies." ], "primary": false, "metadata": { "source": { "provider": "skills.sh", "repository": "https://github.com/yangsonhung/awesome-agent-skills", "path": "skills/en/routerbase-model-routing/SKILL.md", "ref": "HEAD", "url": "https://github.com/yangsonhung/awesome-agent-skills/blob/HEAD/skills/en/routerbase-model-routing/SKILL.md", "key": "yangsonhung/awesome-agent-skills/skills/en/routerbase-model-routing/SKILL.md" } }, "instructions": "# RouterBase Model Routing\n\n## Overview\n\nUse [routerbase](https://routerbase.com/) to plan model selection behind one OpenAI-compatible integration surface. This skill turns workload requirements into a practical model shortlist, fallback chain, cost and latency policy, and validation checklist.\n\n## When to Use\n\nUse this skill when the user asks for:\n- Choosing RouterBase models for chat, coding, reasoning, vision, or multimodal workloads\n- Designing fallback chains for provider outages or degraded quality\n- Balancing cost, latency, context length, output quality, and availability\n- Documentin", "cost": { "context_tokens": 406 } }

routerbase-model-routing

About