Show HN: Galdor – a Go LLM agent framework with built-in tracing and replay (github.com)

Galdor, a Go-native framework for building and orchestrating AI agents, has been released with its first version (v1.0.0). This innovative framework includes built-in tracing and replay capabilities, making it easy to monitor and debug AI agents through an embedded dashboard—eliminating the need for external SaaS solutions. Galdor's standout features are its native OpenTelemetry integration for observability, the ability to self-host the dashboard from a single binary, and support for multiple AI models via a simplified provider interface. The significance of Galdor lies in its unique offering to the AI/ML community, providing a robust self-hostable solution that combines various essential functionalities into one framework. Galdor includes a first-party MCP server, A2A interop support, and an evaluation framework that allows for efficient deployment and management of AI agents. Key technical implications include capabilities for deterministic replay of AI runs, real-time monitoring using OTel spans, and various memory backend support for enhanced agent performance. This release positions Galdor as a strong contender for developers seeking a comprehensive Go-based tool for AI agent orchestration and monitoring.