Assessment instrument · Version 5.1.2
A blank, model-neutral instrument for testing whether a work holds together within its own bounds, carries substantive relations across works or traditions, and supports claims beyond its frame.
Start a new assessment
Use the blank workbook with the exact primary material for a newly declared case. Treat the criteria, protocol, QA definitions, and formula cells as fixed assessment controls.
Declare the medium, tradition-specific conditions, permissions, adaptations, and anything the schema cannot carry.
Use Screening or Standard for bounded, noncertifying work, or Deep for a possible Full assessment. Never present a smaller mode as certification.
Replace the template with a new Case ID and name the exact version, corpus, exclusions, organizing model, and intended claim.
Reconstruct the primary artifact with exact locators, then complete the applicable Internal, Mezo, and External ledgers.
Verify proposed entries, complete Q-001 through Q-040, resolve or disclose disagreements, and report only a scope-bound result.
Automation boundary: AI may extract locators, propose ledger entries, check completeness, and flag candidates. Automated output alone cannot establish a repair ruling, a Deep verdict, or Full qualification.
Three layers
The bounded released artifact must be contradiction-free and substantively organized on its own terms. Peripheral and production evidence may count when it bears on the work.
An unresolved Internal failure blocks qualification. Neither reception nor outside truth can repair it.
At least one substantive relation must be evidenced and typed. The relation records what is inherited, what may be transformed, and which differences carry structural weight.
Resemblance alone is not enough. The relation must do explanatory work.
At least one substantive claim outside the work’s frame must be identified and supported. A work need not be realist to make an External claim.
An unsupported outside claim cannot borrow strength from Internal elegance or Mezo reach.
File boundary
Keeping them separate prevents a convenient finding from rewriting the test and prevents a prior result from becoming an answer key for a new case.
Owns the rules, protocol, evidence standards, blank ledger schemas, formulas, QA definitions, source map, and regression tests. TEMPLATE-001 is an uncompleted scaffold.
Own the populated case inputs, evidence, review, judgments, blocker codes, and results. They may cite an instrument version but cannot alter its criteria.
Full qualification: protocol fit, primary evidence, Deep depth, workbook integrity, all forty QA checks, a qualified conclusion, independent review, and passing Internal, Mezo, and External results are all required. Internal is the absolute gate. Salience, truth-direction, reception, and second-order meaning cannot repair a first-order failure.
Use the instrument
There are no completed case audits in these downloads. Read the Markdown protocol first, then apply the complete blank Excel instrument. The JSON file provides an answer-free machine-readable handoff.
Model-neutral instructions, sheet boundaries, evidence requirements, human decision gates, and output rules
Complete criteria, blank ledgers, formulas, QA controls, sources, and regression tests
Machine-readable file identities, sheet roles, blank case fields, allowed automation, and certification gates
Methodologylist* explains how this assessment connects to the rest of Polymyth. This page is the place to begin or download a Coherence assessment.