active-not-in-antecedents
IN premise
The `active` premise is deliberately excluded from imported beliefs' antecedents; if it were an antecedent, it would provide a second always-valid justification path that defeats per-belief retraction semantics
Summary
When beliefs are imported into the system, the "active" flag is intentionally kept out of their justification chains. If "active" were included as a supporting reason for each belief, every belief would have a backdoor path that keeps it alive regardless of other retractions — effectively making it impossible to retract individual beliefs without deactivating the entire import. This design choice preserves the ability to surgically retract specific beliefs while leaving the rest of the imported set intact.
Dependents
These beliefs depend on this one:
- agent-isolation-through-namespace-and-relay — Agent beliefs are doubly isolated: namespace prefixing prevents ID collisions, while the active/inactive relay pair provides per-agent kill-switch semantics without cross-agent interference
Details
| Source | entries/2026/04/23/topic-multi-agent-federation.md |