AI Skills

Pre-built AI Skills for execution.

Specialized agents that read the context graph, act inside your systems and explain every decision. Configurable per domain, always with a human in the loop.

Filter by family

How Skills work

Configured for your domain. Executed on the context graph. Explained at every step.

01

Configure in Studio

Compose a skill from ontology entities, sources, prompts and policies — with your domain experts, not just engineers.

02

Run on the graph

Skills read the live context graph, call the tools they are authorized to use and write actions back into systems of record.

03

Explain every decision

Every output ships with entities used, signals cited, policies applied and reasoning — ready for review, audit and override.

Studio

Every skill is a template. Open it and make it yours.

Behind every AI Skill is a configurable agent in the Studio. Change the ontology, the sources, the prompt logic, the thresholds and the outputs — without writing code.

01

Configure, not rebuild

Start from a pre-built skill and tune it for your domain. Add entities, signals, policies and review steps in a governed environment.

02

Version and audit

Every change is versioned. Every execution is logged. Every recommendation is traceable back to the model, the data and the reviewer.

03

Human in the loop

Set escalation rules, review gates and override workflows. Compass executes fast; humans stay in control.

Can't find the skill you need? We build it with you.

Every domain has its own decisions. Bring us your workflow and we configure the skill on your context graph.