spans

7 beliefs (2 IN, 5 OUT)

• IN agent-isolation-spans-identity-and-authorization
  Agent beliefs are isolated through two independent containment mechanisms at different system levels: namespace prefixing with relay pairs provides identity-level isolation (preventing ID collisions and enabling per-agent lifecycle control via kill-switch), while transitive subset-gated access tags provide authorization-level isolation (controlling per-caller visibility with tag inheritance).
• IN belief-consistency-spans-structural-and-semantic-dimensions
  The system maintains belief consistency through two complementary mechanisms targeting different quality dimensions: deduplication resolves structural redundancy (merging equivalent beliefs while preserving topology with user-editable auditable plans), while contradiction management resolves semantic inconsistency (traceable dependency-directed backtracking with consistent nogood IDs and minimal-disruption culprit selection).
• OUT data-integrity-spans-architecture
  Data integrity is enforced across all three architectural layers: clean layer boundaries prevent cross-cutting mutations, snapshot persistence ensures atomic state transitions, and conservative staleness checking gates CI pipelines.
• OUT dual-quality-enforcement-spans-automated-and-explicit
  Belief quality is enforced by two independent mechanisms that cannot interfere: automated self-correction autonomously maintains consistency through contradiction resolution and staleness detection, while explicit quality review independently evaluates derived beliefs with read-only fault tolerance — dual enforcement ensures quality even when one mechanism is insufficient.
• OUT edge-case-safety-spans-creation-and-maintenance
  The system handles edge cases safely across both temporal dimensions: at creation time, uniform revision covers all semantic edge cases (vacuous premises, asymmetric absence, empty antecedents) through minimal primitives; at maintenance time, contradiction resolution and staleness detection actively catch drift — no edge case is safe only at one point in time.
• OUT resource-efficiency-spans-footprint-through-budgets
  The system achieves resource efficiency from the broadest to the narrowest scope: zero external dependencies and lazy module loading minimize the static footprint at packaging and startup, while efficient O(1) budget tracking with approximate token estimation constrains resource consumption during both compact distillation and derive belief allocation at runtime.
• OUT resource-efficiency-spans-full-pipeline
  Resource efficiency is enforced across the complete operational pipeline: from packaging and startup (zero external dependencies with lazy loading) through belief derivation (linear O(N) budget allocation with floor bounds) to output generation (O(1) per-line budget tracking with bounded pure compact summaries), ensuring minimal resource consumption at every phase