datasets>=2.0

# tau-bench (benchmark data and task definitions)
tau-bench @ git+https://github.com/sierra-research/tau-bench.git

# tau-bench + OpenAI Agents SDK
openai-agents>=0.10.0
openinference-instrumentation-openai-agents>=1.0.0
arize-phoenix>=13.0.0
opentelemetry-sdk
opentelemetry-exporter-otlp
litellm
openai

# LangGraph (used by both TRAJECT-Bench and tau-bench examples)
langgraph>=0.2
langchain-openai>=0.3
openinference-instrumentation-langchain>=0.1
