Operator-ready prompt for reuse, tuning, and workspace runs.
This item is set up for developers who want to inspect the original language, fork it into Workspace, and adapt the evidence model without losing the source prompt structure.
Implementation handoffs, eval setup, and prompt tuning where you need the original structure intact.
Inspect first, copy once, then fork into Workspace when you want variants, notes, and model settings attached to the same run.
Swap domain facts, examples, and any hard-coded entities for your own context.
Tighten the evidence or verification requirement if this is headed toward production.
Decide which failure mode you want to evaluate first before you branch the prompt.
This prompt already carries implementation detail, tool context, and a final-output instruction. Keep that structure intact when you tune it, or your comparison runs get noisy fast.
Open this prompt inside Workspace when you want a live iteration loop.
Copy for quick reuse, or run it in Workspace to keep prompt variants, model settings, and prompt-history changes in one place.
Structured source with 7 active lines to adapt.
Already linked to a challenge workflow.
Sign in to keep private prompt variations.
Prompt content
Original prompt text with formatting preserved for inspection and clean copy.
To ensure low-latency inference for the autonomous vehicle agents, deploy your GPT-5 Pro and Claude 4 Sonnet instances (or custom fine-tuned models) on Modal and Cerebrium respectively. Then, create custom AutoGen tools that allow your `PathPlanner` and `DecisionMaker` agents to make API calls to these deployed models for their specific reasoning tasks, rather than relying solely on direct `llm_config`. ```python
# Conceptual Modal deployment client
class ModalGPT5ProClient: def plan_path(self, sensor_data: str, current_location: dict, destination: dict) -> str: # Simulate API call to Modal-deployed GPT-5 Pro endpoint return f"Path from Modal GPT-5 Pro for {sensor_data}" # Conceptual Cerebrium deployment client
class CerebriumClaude4SonnetClient: def make_tactical_decision(self, current_situation: str, path_segment: str) -> str: # Simulate API call to Cerebrium-deployed Claude 4 Sonnet endpoint return f"Decision from Cerebrium Claude 4 Sonnet for {current_situation}" # Define AutoGen tools for agents
def modal_path_planning_tool(sensor_data: str, current_location: dict, destination: dict) -> str: """Tool for PathPlanner to query Modal-deployed GPT-5 Pro for path plans.""" return ModalGPT5ProClient().plan_path(sensor_data, current_location, destination) def cerebrium_tactical_decision_tool(current_situation: str, path_segment: str) -> str: """Tool for DecisionMaker to query Cerebrium-deployed Claude 4 Sonnet for tactical decisions.""" return CerebriumClaude4SonnetClient().make_tactical_decision(current_situation, path_segment) # Register tools with agents (example for path_planner)
# path_planner.register_for_llm(name="modal_path_planning", description="...")(modal_path_planning_tool)
```Adaptation plan
Keep the source stable, then branch your edits in a predictable order so the next prompt run is easier to evaluate.
Hold the task contract and output shape stable so generated implementations remain comparable.
Update libraries, interfaces, and environment assumptions to match the stack you actually run.
Test failure handling, edge cases, and any code paths that depend on hidden context or secrets.
Copy once for a pristine source snapshot, then move the prompt into Workspace when you want variants, run history, and side-by-side tuning without losing the original.
Prompt diagnostics
Quick signals for how structured this prompt already is and where adaptation work is likely to happen first.
This prompt already mixes executable detail with instructions, so the safest path is to tune examples and interfaces before you rewrite the overall scaffold.
AutoGen Multi-Agent System for Autonomous Vehicle Simulation & Planning
Drawing inspiration from the emergence of self-driving car companies, this challenge tasks you with developing a robust multi-agent system using AutoGen for simulating autonomous vehicle decision-making and planning. The system will feature specialized agents collaborating to interpret sensor data, generate dynamic path plans, and make real-time tactical decisions in a simulated environment. This project will emphasize the conversational capabilities of AutoGen agents, leveraging GPT-5 Pro for complex strategic planning and Claude 4 Sonnet for nuanced contextual reasoning. Agents will be deployed on Modal and Cerebrium for optimized, low-latency inference, crucial for real-time autonomous systems. An MLflow integration will provide comprehensive MLOps tracking and experimentation capabilities, ensuring robust development and evaluation of the agentic driving policies. The system will simulate a complex urban driving scenario, where agents must coordinate to navigate traffic, react to unexpected events, and adhere to safety protocols.
Use the challenge page to recover the original task boundaries before you tune the prompt. That keeps your variants grounded in the same evaluation target instead of drifting into a different problem.