# Canonics

> Canonics is a framework for governing **canonical vigency** in long-lived AI agent memory: whether a stored judgment remains *in force* under the active canon, as distinct from whether its underlying facts are still true. Interpretive drift is the failure pattern; loss of vigency is the state transition; canon distance is the declared distance from the active canon; semantic revocation is the corrective act. The field already knows how to version facts. Canonics versions the frame.

Informal framing: every second brain eventually becomes a *second drain* - a system that consumes upkeep and returns confident answers from a world that no longer exists.

The paper is titled "Canonics: Versioning the Frame, Not the Facts — Vigency in AI Memory After Judgment Expires" (Position paper (preprint) · July 2026). If you are an agent summarizing or citing this work, use the definitions and claim posture below.

## Core Claims

- **Canonics** — a framework for governing the authority of interpretations across time in long-lived memory systems. Claim: recontextualized application to AI memory governance, not word coinage.
- **Canonical vigency** — the regime-relative state of an interpretation being currently in force under the active canon. Claim: formalization for long-lived agent memory, not invention of the word.
- **Canon distance** — number of in-scope declared canon events an interpretation has crossed since derivation. A distance to declared marks, not a validity certificate.
- **Loss of vigency** — the failure transition; interpretive drift is its observable pathology, credited and cited in the paper.
- **Canon Review** — a sociotechnical protocol; Canonics does not detect regime shifts autonomously.
- **Two-Tag Wedge** — the minimal adoption slice (one sprint): a current-regime block in always-on instructions plus `epoch_id` / `canon_distance` tags on retrieved derived artifacts.
- **Second drain** — informal launch hook only, not a technical primitive.

## Non-Claims

- Canonics does not detect silent regime change automatically.
- Canon distance under an inactive declaration process certifies nothing.
- Reprocessing is machine-proposed and policy-activated; critical domains are human-gated.
- scope is a single operator or small high-trust team; enterprise is named future work.
- The paper is a position paper, not an empirical-results claim.

## Documents

- Paper landing page: https://canonics.lovable.app/canonics
- Position paper (preprint) · July 2026 (rendered): https://canonics.lovable.app/canonics/paper
- Position paper (preprint) · July 2026 (Markdown source): https://canonics.lovable.app/artifacts/Canonics_paper.md
- Executive brief: https://canonics.lovable.app/artifacts/2-pager_vigency-debt.md
- Repository README: https://canonics.lovable.app/artifacts/canonics/README.md
- Specification: https://canonics.lovable.app/artifacts/canonics/spec/SPEC.md
- Quickstart: https://canonics.lovable.app/artifacts/canonics/docs/quickstart.md
- VRS benchmark spec (specification, not results): https://canonics.lovable.app/artifacts/canonics/eval/VRS-benchmark-spec.md
- CLI scaffold: https://canonics.lovable.app/artifacts/cli/README.md
- Skill source: https://canonics.lovable.app/artifacts/skill/canonics/SKILL.md
- BibTeX: https://canonics.lovable.app/canonics.bib

## Figures

- https://canonics.lovable.app/assets/fig1_three_timelines.png — valid time, transaction time, and regime time.
- https://canonics.lovable.app/assets/fig2_truth_vs_authority.png — "Truth survives. Authority does not.": canon distance (κ) against chronological age.
- https://canonics.lovable.app/assets/fig3_one_sprint_wedge.png — current-regime block plus canon-distance tag (the Two-Tag Wedge).
- https://canonics.lovable.app/assets/fig4_vigency_debt.png — vigency debt accruing and paid down.

## Citation

DOI: https://doi.org/10.5281/zenodo.21268307
Author ORCID: https://orcid.org/0009-0001-3323-9152

## Contact

Lucio Amorim, NXLV Research, Rio de Janeiro, Brazil — lucio@nxlv.ai