SkillsTrek: Starship Agency LXP
Immersive workforce learning in a narrative universe, co-created with Trish Uhl. The SkillsTrek Mission Runner coordinates multiple model-driven crew roles through L.O.R.A.
Explore SkillsTrek ↗R&D Lab / Narrative Technology
Systems you can see thinking.
Interactive worlds and working agent experiments offer an unusually useful place to examine orchestration, coherence, and failure in context.
Talk R&D collaboration ↗Research and collaborations
From experiments to evidence.
Selected publicly documented initiatives; partner-reported metrics are credited to their source.
The Disconnected Frontier
A speculative storyworld that also functions as an LLM text-game evaluation sandbox—useful for probing state tracking, narrative continuity and reasoning failures.
Watch TEVV walkthrough ↗TEVV in the open
Practical evaluation walkthroughs with Owl's Ledge put testing, evaluation, verification and validation in view through interactive simulations.
Explore Owl’s Ledge labs ↗The WEASEL Orchestrator
A small, inspectable agent system built around skills, tools, state, stopping rules, execution loops, and traces. Its mechanics inform the Learn Agentic AI book.
See the technical book ↗Owl’s Ledge publishes demonstration metrics for the SkillsTrek Mission Runner. They are collaborator-reported, not independently benchmarked results.
View source ↗Show, don't just tell
Have a system to test or prototype?
Let's discuss the right prototype, simulation or evaluation environment for the question you need answered.
Start an R&D conversation ↗