Prior-art sentinel audit¶

Part of Inverse Innovation with the Encyclopedia of Abstractions · Prior-art sentinel audit · Last revised August 2026

Status: complete post hoc sensitivity audit
Scope: ten hash-selected candidates; forty targeted queries; public-web search only
Interpretive boundary: this is a sentinel, not a population calibration study

Executive result¶

The audit found a closer adjacent precedent in 8 of 10 cases, no closer precedent in 2, and no likely substantial collision with the surviving contrastive claim. All ten original candidates therefore retain their prior-art classification, but eight dossiers should be read with a more densely populated neighboring design space than the first-pass research conveyed.

This is simultaneously reassuring and corrective:

Reassuring: none of the ten candidates collapsed into an already documented equivalent under the audit's strict collision rule.
Corrective: an ADJACENT_PRIOR_ART label does not imply that the original dossier retrieved the strongest historical or operational neighbors.
Limiting: ten deliberately stratified sentinels cannot estimate a corpus-wide false-novelty rate.

Disposition table¶

Candidate	Stratum	Sentinel outcome	What changed
Offline modal-control testing for coupled discovery-exposure imbalance	E3 post hoc	Closer adjacency	Counterfactual offline evaluation makes the evaluation frame less unusual; the modal-control composition remains unfound.
Lifecycle Registry for Accumulated Course Scaffolds	E4 strict	Closer adjacency	Commercial LMS products already encode retirement, replacement, dependencies, and historical completions at course/object level.
Psychometric Norm Lifecycle Registry and Scoring Gate	E5 strict	Closer adjacency	Longstanding testing standards plus scoring platforms cover more of the lifecycle problem than a research-only search suggests.
Label Sunset: Criminal-Record Authority Retirement Cycle	E6 strict	Closer adjacency	Formalized redemption rituals and mature sealing operations are direct conceptual neighbors, but target different actors and outcomes.
Campaign-Bounded Selection for a Film Festival Slate	E6 strict	Closer adjacency	Real festival policies instantiate many package components, though not the full package or its comparative test.
Bounded Rival Reanalysis of Ethnographic Explanations	E6 partner	Closer adjacency	Qualitative restudies and multi-analyst methods explicitly cover rivals and independent reanalysis.
Rotating Proof-Interface Clinic for Cross-Subfield Lemma Handoffs	E6 partner	No closer precedent	Targeted searches returned the same collaboration families already represented in the dossier.
Scenario-Residual Exchange for Distributed Horizon Scanning	E6 partner	Closer adjacency	Ansoff's weak-signal lineage and modern scenario monitoring sharpen the historical baseline.
Representation-Independent Synchronized-Take Contract for Dailies	E6 partner	No closer precedent	IMF/SMPTE/OTIO strengthen the standards baseline but do not instantiate the proposed tested contract.
Why This Control Exists: Control-Memory and Repair Cycle	E6 partner	Closer adjacency	AARs, control self-assessment, and case-based training densely cover components; the ritualized matched comparison remains unfound.

What the audit says about the pipeline¶

The first-pass searches were better at finding recent academic and official sources than at consistently retrieving four kinds of evidence at once: historical lineages, trade/product documentation, non-English material, and full-composition analogues. The audit therefore supports adding explicit search lanes for those four evidence types before a candidate receives a strict novelty-like label.

It does not support changing the report's central claim from “no matching prior art found in the performed search” to “novel.” The search remains bounded, public, and non-expert. Nor does the absence of a substantial collision validate usefulness, feasibility, or effectiveness.

Reproducibility¶

The design was sealed before targeted case searches. Case verdicts were assigned only after comparing new sources with each candidate's original source list and remaining contrastive claim.