0.1.0

BeamWeaver v0.1.0

First public release of BeamWeaver: an OTP-native Elixir library for building AI agents and durable LLM workflows.

This release brings core agentic infrastructure to Elixir so teams can build advanced LLM systems without relying on Python microservices for the orchestration layer.

Highlights

• Agent runtime with tool calling
• Graph workflows with explicit state, routing, and reducers
• Checkpoints and resumable execution
• Short-term memory through checkpoints
• Long-term memory stores
• Retries, fallbacks, interrupts, and human review
• Typed streaming events
• Structured output support
• Provider adapters for OpenAI, Anthropic, Google Gemini, xAI, and Moonshot/Kimi
• Fake and replay models for deterministic tests
• Native tracing hooks, with WeaveScope observability support coming soon