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 23 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.
Initialize a LangGraph state with variables for `interaction_log`, `fluency_behaviors_identified`, `coaching_history`, and `user_feedback`. Define three LangChain agents: an `InteractionAgent` using OpenAI o4-mini to receive user input, a `BehaviorAnalyst` agent using Llama 4 Maverick to interpret interaction logs against the AI Fluency Index, and a `FluencyCoach` agent using OpenAI o4-mini to generate feedback. Ensure each agent has appropriate tools defined for their tasks. ```python
from typing import TypedDict, Annotated, List
from langchain_core.messages import HumanMessage
from langchain_openai import ChatOpenAI
from langchain_community.llms import LlamaCpp
from langgraph.graph import StateGraph, END # Define the state
class AgentState(TypedDict): interaction_log: Annotated[List[str], lambda x, y: x + y] fluency_behaviors_identified: Annotated[List[str], lambda x, y: x + y] coaching_history: Annotated[List[str], lambda x, y: x + y] user_feedback: str # Initialize LLMs
o4_mini_llm = ChatOpenAI(model='GPT-5 Pro-mini', temperature=0.2)
# Placeholder for Llama 4 Maverick, replace with actual loading if available
# Example for LlamaCpp assumes local .gguf file
llama_maverick_llm = LlamaCpp(model_path='/path/to/llama-4-maverick.gguf', temperature=0.1, n_ctx=2048) # Define agents (simplified for prompt, actual implementation would be more complex)
def interaction_agent(state: AgentState): print(f"Interaction Agent received: {state['user_feedback']}") new_log_entry = f"User: {state['user_feedback']}" # Logic to process and potentially generate AI response for next turn return {'interaction_log': [new_log_entry]} def behavior_analyst(state: AgentState): current_interaction = state['interaction_log'][-1] print(f"Behavior Analyst analyzing: {current_interaction}") # Simulate analysis against AI Fluency Index using llama_maverick_llm identified_behaviors = ['Context-setting', 'Iterative clarification'] # Example fluency_score = 7.5 # Example return {'fluency_behaviors_identified': identified_behaviors, 'fluency_score': fluency_score} # Build graph (simplified)
workflow = StateGraph(AgentState)
workflow.add_node('interaction', interaction_agent)
workflow.add_node('analyze', behavior_analyst)
workflow.add_edge('interaction', 'analyze')
# ... more nodes and edges ... # Example of how to add a conditional edge:
# workflow.add_conditional_edges(
# 'analyze',
# lambda state: 'coach' if state['fluency_score'] < 8 else END,
# {'coach': 'coach_node_name'}
# )
```Adaptation plan
Keep the source stable, then branch your edits in a predictable order so the next prompt run is easier to evaluate.
Preserve the role framing, objective, and reporting structure so comparison runs stay coherent.
Swap in your own domain constraints, anomaly thresholds, and examples before you branch variants.
Check whether the prompt asks for the right evidence, confidence signal, and escalation path.
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.
AI Fluency Index Evaluator with LangGraph and OpenAI o4-mini
Anthropic's AI Fluency Index highlights key behaviors for effective human-AI collaboration. This challenge involves building a multi-agent system using LangChain with LangGraph to act as an "AI Fluency Coach." The system will interact with users, observe their collaboration patterns (simulated), evaluate these against the Fluency Index behaviors, and provide actionable feedback. It will utilize specialist agents powered by OpenAI o4-mini and Llama 4 Maverick for understanding user input, analyzing behavior, and generating coaching advice. The objective is to demonstrate how graph-based agent orchestration can create dynamic, adaptive evaluation and improvement systems for human-AI interaction.
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.