Which model is best for daily multi-file debugging?

Sonnet is the recommended choice for 80-90% of daily coding tasks, including multi-file reasoning and standard debugging.

When should I choose Haiku over Sonnet or Opus?

Haiku is best for well-defined, high-volume tasks with detailed prompts where speed and cost-efficiency are more important than complex decision-making.

Is Opus worth the higher cost for coding?

Opus is recommended for high-stakes decisions and complex analysis where sustained focus and nuanced judgment are required to prevent looping.

How do I prevent Sonnet from getting off-track?

When using Sonnet, use specific prompting to guard against its tendency to explain concepts or gather extraneous info instead of executing the task.

Can I manually override the agent selection logic?

Yes, operator directions always supersede the skill's recommendations; if you specify a model, the system will execute with that agent.

AI Agent Selection Strategy

Name: AI Agent Selection Strategy
Author: ed3dai

byed3dai

•

Productivity & Workflow

Optimizes developer workflows by intelligently selecting the most efficient Claude model for specific coding tasks based on complexity and cost.

The using-generic-agents skill provides a strategic framework for choosing between Claude model variants—Haiku, Sonnet, and Opus—to balance operational costs, execution speed, and reasoning depth. By providing clear heuristics for model selection, it ensures that high-volume, well-defined tasks are handled by Haiku, while complex, high-stakes architectural decisions are reserved for Opus, effectively preventing common pitfalls like model wandering or inefficient token usage.

Key Features

01Cost-optimization heuristics for parallel workflows

02Heuristics for multi-file reasoning and debugging

03Strategic model-to-task mapping for Haiku, Sonnet, and Opus

04Guardrails to prevent model distraction and looping

05Specific prompting guidance based on model characteristics

0645 GitHub stars

Use Cases

01Selecting Haiku for high-volume, parallelized execution of simple code updates

02Deploying Opus for high-stakes decisions and complex logic analysis where focus is critical

03Utilizing Sonnet for standard daily coding tasks and multi-file debugging

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ed3dai/ed3d-plugins using-generic-agents

For use in Claude.ai and ChatGPT

Download Skill