Kapixhint is designed to function as an explanatory model and reading aid for budget documents. The model articulates clear, neutral descriptions of categories, constraints, and recorded changes so that observers can interpret structure and lineage without prescriptive commentary. It is intended for reference use where users seek to understand how a budget is assembled, how categories relate, and how recorded adjustments are expressed. The scope explicitly excludes procedural or operational instructions; it does not advise action or decision-making. Instead, it records taxonomy, boundary definitions, and revision histories that make the document state auditable and comparable across versions. The structure supports hierarchical and networked category mappings, conditional qualifiers for boundaries, and typed revision markers that preserve minimal but sufficient metadata for interpretation. Documentation emphasizes consistent label schemes, provenance lines, and optional example scenarios that illustrate edge cases. The reference model is suitable for archival reading, cross-document comparison, and explanatory annotations that accompany ledger-style excerpts.

The model defines a small set of primitives that together describe a budget document. Category is the primary labeling primitive and includes a stable identifier, a human-readable label, and an explicit scope statement that enumerates inclusion criteria. Boundary is a constraint primitive that records limiting dimensions, measurement units, and activation conditions; boundaries are typed as hard or soft and include an origin reference with an effective date. Revision marker is a change primitive describing deltas, replacement values, or reclassifications; each marker contains a concise summary, affected category references, a timestamp, and a rationale field. Contextual note is an auxiliary primitive used to capture assumptions, sampling windows, measurement methods, or uncertainty ranges; notes are anchored to a specific location and can be toggled between terse and expanded views. Aggregation rules document how subcategory values contribute to higher-level aggregates and record derivation formulas and sample dates. Each primitive includes minimal provenance metadata (author identifier, timestamp, and optional document identifier) to support traceable reading. The primitives are designed to be machine-readable and human-auditable, facilitating neutral interpretation without implying procedural intent.

The provenance model records the chain of custody for labels, values, and boundary statements. Each revision entry preserves a clear link to any prior version it supersedes and records the type of change: annotation, adjustment, reclassification, or correction. A revision entry lists affected primitives and provides a minimal rationale and timestamp. When revisions impact aggregated values, the provenance entry includes the propagation rule used to recompute aggregates and identifies the exact derivation formula where applicable. Provenance lines include an author identifier and an optional external document reference to enable verification against supplementary materials. The model supports filtered views of revision histories by type, date range, or affected category to reduce cognitive load when inspecting change sets. Visual presentation of revision markers uses muted stamps and inline micro-notes so readers can scan for context without conflating unrelated changes. The objective is to enable neutral comprehension of how a document evolved while preserving sufficient metadata for auditability and reproducibility of derived values.