Implement a capability-based rubric for model selection and task routing.
npx skills add https://github.com/AesopScott/mojo --skill selection-rubric
Please help me install this Agent Skill. Skill: selection-rubric Source: https://github.com/AesopScott/mojo/tree/main/harnesses/skills/selection-rubric Command: npx skills add https://github.com/AesopScott/mojo --skill selection-rubric
This Skill resolves ambiguity in model routing by providing a structured rubric to map specific task types to the most appropriate AI model based on capability, cost, and latency requirements.
Use the selection-rubric skill to audit the current model routing logic and propose a new capability mapping for the upcoming task type.
High-intent search queries and answers about installing and using this skill.
Capability-based model selection is a structured approach that maps specific task types to the most appropriate AI model by evaluating capability, cost, and latency requirements to prevent model mismatch and over-provisioning.
To optimize model routing, define the required capability threshold for the new task type using a structured rubric, ensuring the system selects a cost-effective model that meets performance criteria without exceeding authority.
Over-provisioning is prevented by implementing controlled routing with a capability catalog that defines clear boundaries for what each model tier can handle, ensuring objective-based evaluation criteria are met.
You need defined capability catalogs and objective-based evaluation criteria established in your system architecture and harness configuration to ensure consistent, secure, and strictly enforced model assignment boundaries.
A rubric-based decision process should be used when strict governance is required to balance model performance, cost, and safety boundaries across varying task-specific needs in your system architecture.
az9713Provide prompting patterns, failure modes, and model-specific guidance for LLM workflows.
nanocoaiConfigure NanoClaw agent groups to use local Ollama models via environment variables.
daemon-blockint-techGuides ROS2-based autonomous systems to operate within latency and safety budgets.
latestaiagentsRoute LLM tasks to models based on complexity, cost, and latency.
