Wiki / 6 of 7
How the parts work together
EolSpace connects intent, work, evidence, permission, outcome and versions. The surrounding roles stay distinct; the older Sun, planets and satellites model is historical.
EolSpace at the centre
A person defines the goal and its limits. In the intended system, EolSpace keeps the links between completed work, its evidence, an authorized decision and the observed outcome.
An unknown does not become success because a report sounds confident. Evidence for an earlier version does not establish the new version automatically.
Surrounding roles
People and AI agents prepare results. A contributor does not acquire the authority to approve their own work merely by completing it.
Checks establish specific properties under stated conditions. JEV may be a semantic tool, but its assessment is not permission to act.
TypeSafeAI separately studies the reliability of tool assessments. EolSpace defines their permitted use; this page does not claim a live runtime connection.
Historical model
The Sun, planets and satellites described an internal work topology in the 26 August snapshot. This is a historical record, not the active public topology of EolSpace.
The old names and roles are not mandatory components of the new product. One useful principle remains: a promise, completed work and evidence mean different things.
Continuation
The observed outcome returns to the work history and informs the next version without rewriting the initial goal.