Name: model
Author: xiaobei930

Imported from xiaobei930/cc-best (skills/model/SKILL.md). Install upstream with npx skills add xiaobei930/cc-best --skill model. Copyright stays with the author.

模型选择技能

本技能提供任务类型到模型的推荐映射，帮助 Claude 在不同场景下选择合适的模型。

子文件

routing-matrix.md - 任务类型→模型推荐矩阵

核心原则

探索用 Haiku — 快速、低成本，适合大量 Glob/Grep/Read 操作
实现用 Sonnet — 平衡质量和速度，适合编码和测试
决策用 Opus — 深度思考，适合架构设计和安全审查

与 /cc-best:model 命令的关系

/cc-best:model 命令用于全局切换所有 Agent 的模型策略（quality/balanced/economy）
本技能提供细粒度参考，帮助 Claude 在具体任务中判断模型选择
命令是"批量切换"，技能是"逐任务建议"

使用场景

自动参考

当 Claude 遇到以下场景时，会自动参考本技能：

启动子代理（Task tool）时选择 model 参数
评估当前任务复杂度时
用户询问模型选择建议时

决策流程

任务到达 ──→ 评估复杂度 ──→ 查询 routing-matrix ──→ 选择模型
                │
                ├─ 简单（搜索/读取）→ haiku
                ├─ 中等（编码/测试）→ sonnet
                └─ 复杂（架构/安全）→ opus

最佳实践

成本敏感场景 — 子代理探索任务优先 haiku，节省 token
质量敏感场景 — 安全审查、架构决策必须用 opus
混合场景 — 同一工作流中可对不同阶段使用不同模型
用户偏好 — 尊重 /cc-best:model 设定的全局策略

记住：模型选择的目标是用最低成本达到足够的质量。不是所有任务都需要最强模型，也不是所有任务都能用最快模型。

# after Install: the listing is in your workspace's routing pool - a plan picks it for its slot curl -s -X POST https://api.opensmartroute.ai/api/v1/route -H 'Authorization: Bearer $OSR_API_KEY' -H 'Content-Type: application/json' -d '{"text": "...", "plan": true}'

Install model 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": "xiaobei930-cc-best-model", "kind": "skill", "name": "model", "description": "Model selection strategy and routing. Use when choosing between models for different task types, subagent configurations, or optimizing token cost vs quality tradeoffs.", "publisher": "xiaobei930", "version": "1.0.0", "capabilities": { "domains": [ "general" ], "tags": [ "skill-md", "github" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "Model selection strategy and routing. Use when choosing between models for different task types, subagent configurations, or optimizing token cost vs quality tradeoffs." ], "primary": false, "metadata": { "source": { "provider": "github", "repository": "https://github.com/xiaobei930/cc-best", "path": "skills/model/SKILL.md", "ref": "a75ef2595230ad53fdd1711fd0cb6250d833541c", "url": "https://github.com/xiaobei930/cc-best/blob/a75ef2595230ad53fdd1711fd0cb6250d833541c/skills/model/SKILL.md", "key": "xiaobei930/cc-best/skills/model/SKILL.md" }, "allowed_tools": [ "Read,", "Grep,", "Glob" ] }, "instructions": "# 模型选择技能\n\n本技能提供任务类型到模型的推荐映射，帮助 Claude 在不同场景下选择合适的模型。\n\n## 子文件\n\n- [routing-matrix.md](routing-matrix.md) - 任务类型→模型推荐矩阵\n\n## 核心原则\n\n1. **探索用 Haiku** — 快速、低成本，适合大量 Glob/Grep/Read 操作\n2. **实现用 Sonnet** — 平衡质量和速度，适合编码和测试\n3. **决策用 Opus** — 深度思考，适合架构设计和安全审查\n\n## 与 /cc-best:model 命令的关系\n\n- `/cc-best:model` 命令用于**全局切换**所有 Agent 的模型策略（quality/balanced/economy）\n- 本技能提供**细粒度参考**，帮助 Claude 在具体任务中判断模型选择\n- 命令是\"批量切换\"，技能是\"逐任务建议\"\n\n## 使用场景\n\n### 自动参考\n\n当 Claude 遇到以下场景时，会自动参考本技能：\n\n- 启动子代理（Task tool）时选择 model 参数\n- 评估当前任务复杂度时\n- 用户询问模型选择建议时\n\n### 决策流程\n\n```\n任务到达 ──→ 评估复杂度 ──→ 查询 routing-matrix ──→ 选择模型\n │\n ", "cost": { "context_tokens": 233 } }

model

About

模型选择技能

子文件

核心原则

与 /cc-best:model 命令的关系

使用场景

自动参考

决策流程

最佳实践

Use it

Install model and it becomes one more candidate the router can pick - when it fits.

Manifest

Reviews