skills.rest
Arm-Examples
Arm-ExamplesOFFICIAL
Explore all skills by this author →
GitHub

llm-add-model-support

Add and revise LLM model configuration JSON files with CTest integration.

4|3|Updated Feb 18, 2026
Search Tags:#testing#llm#configuration#cmake#backend integration#model onboarding
One-click install
npx skills add https://github.com/Arm-Examples/LLM-Runner --skill llm-add-model-support
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: llm-add-model-support
Source: https://github.com/Arm-Examples/LLM-Runner/tree/main/skills/llm-add-model-support
Command: npx skills add https://github.com/Arm-Examples/LLM-Runner --skill llm-add-model-support

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill streamlines the process of adding or updating model configurations and ensuring they are properly tested within the LLM-Runner framework, simplifying model integration and maintenance.

Core Features & Use Cases

  • Model Configuration: Add or revise JSON configurations for LLM models.
  • Test Integration: Ensure new or updated models are exercised by CTest.
  • End-to-End Onboarding: Optionally manage model asset downloads, backend quirks, and documentation updates.
  • Use Case: When a new open-source LLM is released, use this Skill to add its configuration, download its weights, and verify it runs correctly with the LLM-Runner's test suite.

Quick Start

Add a new model configuration JSON for the 'llama-2-7b' model under the 'llama.cpp' backend.

Frequently Asked Questions about llm-add-model-support

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

FAQPage Schema
How do I add a new LLM model configuration to an existing backend framework?▼

Adding a new LLM model configuration involves creating or revising a JSON file tailored to your specific backend, adjusting prompts and templates as needed. This ensures the model integrates seamlessly and is ready for comprehensive testing within the LLM-Runner framework.

Does CTest integration support testing new LLM model configurations automatically?▼

Yes, CTest integration ensures new or updated LLM models are exercised within the test suite. This validates that the model configuration JSON files and backend integrations function correctly after the onboarding process.

Can I use this to onboard models for onnxruntime-genai and mediapipe frameworks?▼

Yes, model onboarding supports llama.cpp, onnxruntime-genai, mediapipe, and MNN frameworks. You can manage backend-specific prompt adjustments and template configurations for any of these supported environments.

What is the process for managing LLM model asset downloads during onboarding?▼

The onboarding process optionally manages end-to-end model asset downloads alongside backend-specific adjustments. This handles acquiring necessary weights and assets, verifying the model runs correctly with the test suite.

How do I update documentation when revising LLM model configurations?▼

Updating documentation is managed as part of the end-to-end model onboarding process. When you revise a model configuration JSON, the workflow ensures corresponding documentation updates are generated to reflect the new backend integrations.

Why do I need to adjust backend-specific prompts when adding an LLM model?▼

Backend-specific prompt adjustments are necessary because frameworks like llama.cpp and onnxruntime-genai require unique templates. Adjusting these ensures the LLM model interprets inputs correctly and functions properly within its specific backend.

Related Skills

Explore More Software Engineering Skills→
heyAyushhheyAyushh
3

onboard-new-developer

Configure development environments and prepare initial pull requests for new developers.

Community
Basic
Arm-ExamplesArm-Examples
4

llm-debug-test-failures

Debug failing LLM integration tests with verbose ctest reruns and Python scripts.

Official
Intermediate
cyberpunk042cyberpunk042

infra-cache

Configure per-model prompt and KV cache settings in LocalAI YAML files.

Community
Intermediate
khokskhoks
2

user_custom_strategy_intake

Scaffold custom trading strategy directories and register them in ClaudeTrading configuration.

Community
Intermediate