Brian Anderson releases ASE, VulcanAMI, FEMS

This reads less like a polished product launch than a raw systems drop, but that’s also what makes it interesting: the code appears to have real deployment surface area and serious infra behind it, even if the finish is uneven.

• –ASE targets autonomous software creation and feedback loops, putting it squarely in the agentic coding/tooling lane.
• –VulcanAMI is the broadest swing: a hybrid symbolic/neural AI stack with governance, observability, and self-improvement claims layered on top.
• –FEMS focuses on counterfactual and scenario simulation, which makes it more of a scientific/infrastructure play than a consumer-facing app.
• –The common thread is operational maturity, not polish: Docker, Helm, Kubernetes, Prometheus, and Grafana show up repeatedly, so this is built to run, not just demo.
• –The biggest risk is validation. The release itself admits bugs, incomplete docs, and security gaps, so outside contributors should treat the repos as foundations, not benchmarks.