skills.rest
Izak90
Izak90COMMUNITY
Explore all skills by this author →
GitHub

Model Selection

Resolve LLM models for agent spawns using a 5-layer hierarchy.

Updated Nov 27, 2024
Search Tags:#llm#fallback#model-selection#charter#config.json#agent-spawn
One-click install
npx skills add https://github.com/Izak90/Izak90.github.io --skill model-selection-izak90
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: Model Selection
Source: https://github.com/Izak90/Izak90.github.io/tree/main/isaac_treinos_ar_livre/.copilot/skills/model-selection
Command: npx skills add https://github.com/Izak90/Izak90.github.io --skill model-selection-izak90

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Determines and enforces the correct LLM model for each agent spawn by applying a deterministic 5-layer resolution hierarchy, balancing per-agent overrides, global defaults, session directives, charter preferences, and task-type considerations.

Core Features & Use Cases

  • Persistent model preferences in .squad/config.json and per-agent overrides
  • Layered resolution: Per-Agent Config, Global Config, Session Directive, Charter, Task-Aware Auto, with a defined fallback
  • Spawn acknowledgments include the resolved model for traceability
  • Supports updating and clearing model preferences to control automatic selection

Quick Start

Resolve and apply the appropriate model for an agent spawn using the defined 5-layer hierarchy.

Frequently Asked Questions about Model Selection

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I assign the right LLM model for each agent spawn?▼

Assign the right LLM model for each agent spawn by applying a deterministic 5-layer resolution hierarchy that checks per-agent overrides, global defaults, session directives, charter preferences, and task type. This resolves the appropriate model automatically with defined fallbacks.

How does the 5-layer model selection hierarchy work for agents?▼

The 5-layer model selection hierarchy resolves the correct LLM by sequentially checking per-agent config, global config, session directives, charter preferences, and task-aware auto-selection. It ensures deterministic fallbacks if higher layers lack defined preferences.

Can I set persistent LLM preferences for agents in config.json?▼

Yes, you can set persistent LLM preferences in .squad/config.json. The system stores global defaults and per-agent overrides there, allowing you to update or clear model preferences to control automatic selection across spawn events.

What is the best way to override a global LLM default for a specific agent?▼

The best way to override a global LLM default is to use per-agent configuration. The resolution hierarchy prioritizes per-agent overrides above global config, session directives, and charter preferences to ensure the specific agent uses the intended model.

How do I verify which LLM model was assigned during an agent spawn?▼

Verify the assigned LLM model by checking the spawn acknowledgment. The system includes the resolved model in each spawn acknowledgment output, providing full traceability for which LLM was selected through the 5-layer hierarchy.

Why does my agent spawn use a fallback model instead of the configured LLM?▼

An agent spawn uses a fallback model when no valid preference is found across the 5-layer hierarchy. If per-agent overrides, global config, session directives, charter preferences, and task-aware auto-selection all lack a defined model, the deterministic fallback is applied.

Related Skills

Explore More Software Engineering Skills→
lbouriezlbouriez

Model Selection

Determines the LLM model for each agent spawn using a layered priority approach.

Community
Advanced
quaid-appquaid-app
3

Model Selection

Resolve and persist LLM model choices per agent spawn in .squad/config.json.

Official
Advanced
dotnetdotnet
209

Model Selection

Resolve LLM models per agent spawn via layered first-match-wins hierarchy.

Official
Intermediate
DeDuvaDeDuva
1

model-selection

Assign LLM models to agent spawns using a multi-layer resolution hierarchy.

Community
Advanced