AI

Common ground for many agents—not one central intelligence.

Ministry agents can widen access to knowledge and service. They can also centralize authority, conceal sources, amplify error, and extract value from communities. Shared protocols make those choices visible.

A trustworthy agent declaration

What people deserve to know.

Identity and purpose

Who operates the agent, what it is intended to do, and who is accountable for it.

Knowledge provenance

Sources, licensing, material limitations, update dates, and how contributors are credited.

Values and evaluation

Declared theological and ethical commitments, test methods, results, known failure modes, and review history.

Tools and privacy

What data the agent receives, retains, or shares; what actions it can take; and when human approval is required.

Interoperability

Capabilities should be able to work together.

Builders should be able to contribute a translation workflow, evaluation set, research corpus, safety review, interface, or specialist tool without surrendering everything to one owner.

Common declarations, permission boundaries, evaluation evidence, audit history, and contribution records can help people discover and combine trusted capabilities while allowing different ownership and licensing models.

Fair return

Contribution is broader than code.

Theologians, translators, educators, local communities, testers, moderators, designers, data stewards, and people who document failures all create value. Empty Jars will explore ways to preserve attribution and return recognition or financial value without claiming that an algorithm can perfectly measure contribution.

Register as an agent builder