Changelog

All notable changes to TraceCraft will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

Unreleased

Initial release of TraceCraft
Core instrumentation decorators (@trace_agent, @trace_tool, @trace_llm, @trace_retrieval)
Console and JSONL exporters
OTLP exporter for OpenTelemetry Protocol support
PII redaction processor with built-in patterns
Sampling processor for trace volume control
LangChain integration via TraceCraftCallbackHandler
LlamaIndex integration via TraceCraftSpanHandler
PydanticAI adapter
Claude SDK wrapper
Auto-instrumentation for OpenAI and Anthropic SDKs
Terminal UI (TUI) for interactive trace exploration
Multi-tenant runtime support
AWS AgentCore integration
Azure AI Foundry integration
GCP Vertex Agent Builder integration
MLflow exporter
HTML report exporter
Comprehensive test suite
Documentation and examples

TraceCraft is now available as a vendor-neutral LLM observability SDK. Key features:

Local-First Development - Beautiful console output and JSONL files without any backend
Built on OpenTelemetry - Industry-standard foundation
Framework Agnostic - Works with LangChain, LlamaIndex, PydanticAI, and custom code
Privacy by Default - PII redaction built into the SDK
Multiple Export Targets - Console, JSONL, OTLP, MLflow, HTML

Install with:

pip install tracecraft

See the documentation for complete usage guide.