home / skills / hexbee / hello-skills / china-model-selection-guide

china-model-selection-guide skill

/skills/china-model-selection-guide

This skill helps select the best-fit model from several options based on input type, task complexity, and delivery goals to optimize routing.

npx playbooks add skill hexbee/hello-skills --skill china-model-selection-guide

Review the files below or copy the command above to add this skill to your agents.

Files (4)
SKILL.md
2.3 KB
---
name: china-model-selection-guide
description: China model selection and task-routing guide for Doubao-Seed-2.0-Code, GLM-5, MiniMax-M2.5, and Kimi-K2.5. Use when users need to choose the best-fit model by input type, task complexity, engineering constraints, and delivery goals, including staged multi-model workflows.
---

# China Model Selection Guide

Follow this flow to recommend models. Load `references/china-model-selection-guide.md` for the full Chinese playbook, scenarios, strengths, and prompt templates.

## Quick Triage

Answer two questions first, then give a primary pick.

1. Identify core input type
- Visual-first input (UI mockups, screenshots, sketches): prefer `Doubao-Seed-2.0-Code`
- Very long text or many files (dozens of docs, full codebase): prefer `Kimi-K2.5`
- Structured engineering prompts (clear coding requirements, Shell commands): prefer `GLM-5` or `MiniMax-M2.5`

2. Identify task complexity
- Complex reasoning or autonomous planning (system design, codebase refactor): prefer `GLM-5`
- Cross-language engineering (Python/C++, Java/Go): prefer `MiniMax-M2.5`
- Clear task but heavy execution (UI-to-code, template generation): prefer `Doubao-Seed-2.0-Code`

## Tie-Break Rules

When multiple models fit, decide in this order.

1. Satisfy hard constraints first: vision, long-context, cross-language, agentic planning
2. Then compare cost and latency: pick better price/performance at similar quality
3. Finally split by phase: allow multi-model routing inside one project

## Composite Task Routing

Use this default pipeline.

1. Planning: `GLM-5` for architecture, decomposition, interfaces, schema decisions
2. Build:
- Frontend and visual replication: `Doubao-Seed-2.0-Code`
- Backend scripts, cross-language tasks, terminal automation: `MiniMax-M2.5`
3. Integration debugging: route hard cross-module issues back to `GLM-5`
4. Documentation handoff: send codebase and large document sets to `Kimi-K2.5`

## Output Format

Always include these in recommendations.

1. Decision: primary model + fallback model
2. Rationale: map to input type, complexity, and constraints
3. Risks: likely weak points and rollback strategy
4. Execution: a ready-to-use prompt draft

## References

- Full guide and examples (Chinese): `references/china-model-selection-guide.md`

Overview

This skill provides a compact China-focused model selection and task-routing guide for Doubao-Seed-2.0-Code, GLM-5, MiniMax-M2.5, and Kimi-K2.5. It helps teams pick the best-fit model based on input type, task complexity, engineering constraints, and delivery goals. The skill also defines a default multi-model pipeline for staged projects and mandates a consistent recommendation output format.

How this skill works

The guide first triages by input type (visual, long-text, structured prompts) and task complexity (reasoning, cross-language, execution-heavy). It applies tie-break rules that prioritize hard constraints (vision, long context, cross-language, agentic planning), then cost/latency, then project phase to select primary and fallback models. For composite tasks it prescribes a default pipeline: GLM-5 for planning, Doubao for visual/frontend, MiniMax for backend/automation, and Kimi for large-document handoff. Every recommendation includes decision, rationale, risks, and an execution-ready prompt draft.

When to use it

  • Choosing a single model for a new engineering or product task
  • Designing a staged pipeline that uses multiple models per phase
  • Routing tasks according to input modality (images, long text, code)
  • Balancing quality versus latency and cost under engineering constraints
  • Preparing handoff prompts and rollback strategies for production

Best practices

  • Start by answering two triage questions: core input type and task complexity
  • Prioritize hard constraints (vision, long-context, cross-language) before cost
  • Use GLM-5 for architecture, decomposition, and agentic planning
  • Route visual/frontend work to Doubao-Seed-2.0-Code and cross-language code to MiniMax-M2.5
  • Send large document sets or full codebase handoffs to Kimi-K2.5 for downstream tasks

Example use cases

  • Build a UI mockup-to-code pipeline: Doubao primary, MiniMax fallback for integration scripts
  • Refactor a multilingual codebase: plan with GLM-5, implement cross-language patches with MiniMax
  • Process dozens of technical docs for release notes and API docs: Kimi-K2.5 primary
  • Complex system design and agent planning: GLM-5 primary with Kimi for documentation handoff
  • Multi-phase product delivery: planning (GLM-5) → frontend (Doubao) → backend (MiniMax) → docs (Kimi)

FAQ

What if multiple models match the triage?

Follow tie-break rules: satisfy hard constraints first, then compare cost/latency, and finally split by project phase to allow multi-model routing.

How should recommendations be presented?

Always include: Decision (primary + fallback), Rationale mapped to input/constraints, Risks and rollback strategy, and an execution-ready prompt draft.