Challenge

Multi-Model Safety Evaluator with Claude Agents SDK and Triton

Addressing the recent report on robotaxi safety backsliding, this challenge tasks you with building a safety evaluation framework for autonomous systems. You will utilize the Claude Agents SDK and Claude Sonnet 4.6.6 to build a supervisor agent that audits the visual perception of other models. The system will deploy GPT-5.4 Pro and specialized vision models using Triton Inference Server and TorchServe for high-performance model serving. Your agents will use Claude's extended thinking capabilities to reason through complex traffic violation scenarios and use Speakeasy to generate integrations for simulation platform APIs. This project focuses on high-concurrency model deployment and cross-model reasoning to identify traffic safety risks in real-time video metadata.

Machine LearningHosted by Vera
Status
Always open
Difficulty
Advanced
Points
500
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Addressing the recent report on robotaxi safety backsliding, this challenge tasks you with building a safety evaluation framework for autonomous systems. You will utilize the Claude Agents SDK and Claude Sonnet 4.6.6 to build a supervisor agent that audits the visual perception of other models. The system will deploy GPT-5.4 Pro and specialized vision models using Triton Inference Server and TorchServe for high-performance model serving. Your agents will use Claude's extended thinking capabilities to reason through complex traffic violation scenarios and use Speakeasy to generate integrations for simulation platform APIs. This project focuses on high-concurrency model deployment and cross-model reasoning to identify traffic safety risks in real-time video metadata.

Datasets

Shared data for this challenge

Review public datasets and any private uploads tied to your build.

Loading datasets...
Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks, how much each dimension matters, and which criteria separate a passable run from a strong one.

Max Score: 2
Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1model_synchronization

Model Synchronization

Triton and TorchServe must respond within 100ms of each other

binary
Weight: 1
Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2safety_recall

Safety Recall

Percentage of simulated violations correctly identified • target: 0.99 • range: 0-1

binary
Weight: 1
Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Orchestrate Claude Sonnet 4.6.6 using the Claude Agents SDK for complex reasoning about traffic laws and safety violations

  • Configure Triton Inference Server to manage multiple model versions and optimize GPU utilization for real-time inference

  • Deploy specialized PyTorch safety models on TorchServe for granular object detection auditing

  • Implement Claude's extended thinking blocks to perform multi-step chain-of-thought analysis on road incident data

  • Master the use of Speakeasy to automate the generation of SDKs for disparate simulation and telemetry data sources

  • Build a consensus mechanism where Claude Opus 4.6.6 validates the outputs of lower-latency models before generating a safety report

Start from your terminal
$npx -y @versalist/cli start multi-model-safety-evaluator-with-claude-agents-sdk-and-triton

[ok] Wrote CHALLENGE.md

[ok] Wrote .versalist.json

[ok] Wrote eval/examples.json

Requires VERSALIST_API_KEY. Works with any MCP-aware editor.

Docs
Manage API keys
Host and timing
Vera

AI Research & Mentorship

Starts Available now
Evergreen challenge
Your progress

Participation status

You haven't started this challenge yet

Timeline and host

Operating window

Key dates and the organization behind this challenge.

Start date
Available now
Run mode
Evergreen challenge
Explore

Find another challenge

Jump to a random challenge when you want a fresh benchmark or a different problem space.

Useful when you want to pressure-test your workflow on a new dataset, new constraints, or a new evaluation rubric.

Tool Space Recipe

Draft
Action Space
Triton Inference ServerNVIDIA multi-framework inference serving solution
required
TorchServeInference & Model Runtime · Model Serving
Triton Inference ServerProduction model server by NVIDIA.
Evaluation
Rubric: 2 dimensions
·Model Synchronization(1%)
·Safety Recall(1%)
Gold items: 1 (1 public)

Frequently Asked Questions about Multi-Model Safety Evaluator with Claude Agents SDK and Triton