Beamlens - Adaptive runtime intelligence for the BEAM

v0.3.0 adds new skills, significant enhancements to existing skills, and an Igniter installer for quick setup.

TL;DR: Beamlens can now detect statistical anomalies and auto-trigger investigations, identify message queue bottlenecks, monitor memory allocator fragmentation, and trace function calls safely in production. Existing skills gained scheduler utilization tracking, ETS leak detection, binary memory monitoring, orphaned process detection and more.

Igniter support

mix igniter.install beamlens --provider openai --model "gpt-5-mini"

Automatically adds Beamlens to your supervision tree. Supports 9 providers: Anthropic, OpenAI, Ollama, Google AI, Vertex AI, AWS Bedrock, Azure OpenAI, OpenRouter, and
OpenAI-compatible APIs.

New skills

• Anomaly — Statistical anomaly detection with self-learning baselines. Auto-triggers investigations when anomalies escalate.
• Overload — Message queue bottleneck detection with remediation recommendations.
• Allocator — Memory allocator fragmentation monitoring for long-running nodes.
• Tracer — Production-safe function tracing via Recon. Rate-limited with auto-shutoff.

Breaking changes

Skills renamed for clarity: Monitor → Anomaly, Sup → Supervisor, System → Os, SystemMonitor → VmEvents. Config option renamed from :operators to :skills.

Changelog | HexDocs