CM Capability survey invariants
metadata
| Title: | CM-1 Capability survey invariants |
| Author: | Ralph B. Holland |
| Publication date | 2026-01-04T14:53Z |
| Affiliation | Arising Technology Systems Pty Ltd |
| Contact | ralph.b.holland [at] gmail.com |
| Provenance | Authored MediaWiki artefact; edit history reflects editorial changes only |
| Binding | Normative (verbatim) |
Metadata (Normative)
The metadata table immediately preceding this section is CM-defined and constitutes the authoritative provenance record for this MWDUMP artefact.
All fields in that table (including artefact, author, version, date, local timezone, and reason) MUST be treated as normative metadata.
The assisting system MUST NOT infer, normalise, reinterpret, duplicate, or rewrite these fields. If any field is missing, unclear, or later superseded, the change MUST be made explicitly by the human and recorded via version update, not inferred.
Curator Provenance and Licensing Notice
This document predates its open licensing.
As curator and author, I apply the Apache License, Version 2.0, at publication to permit reuse and implementation while preventing enclosure or patent capture. This licensing action does not revise, reinterpret, or supersede any normative content herein.
Authority remains explicitly human; no implementation, system, or platform may assert epistemic authority by virtue of this license.
Applicability Note
The invariants and predicates defined in this section SHALL be applied to all market survey executions from Version 2 onward.
They SHALL NOT be used to reinterpret, correct, or invalidate Version 1 results, which remain a valid historical snapshot executed under the definitions and evidence available at that time.
---
I. Capability Definition Invariants (Authoritative)
All capability meanings are fixed by: Cognitive Memoisation: LLM Systems Requirements for Knowledge Round Trip Engineering.
No platform documentation, analyst interpretation, or inferred behaviour SHALL override these definitions.
The following requirements are locked:
- Semantic Artefact Ingress
- Durable Artefact Egress
- Context Determinism
- Transport Independence
- Governance Compatibility
- On-Site / Embedded Suitability (deployment classification only)
Symbol semantics are fixed as:
- ✓ (satisfies requirement)
- ✗ (fails requirement)
- ? (insufficient evidence to determine)
Successful CM round-trip (emit → capture → re-ingest → assert → govern) SHALL be treated as decisive evidence for:
- ingress,
- egress,
- determinism,
- and transport independence.
---
II. Survey Execution Invariants
- Each platform surface (Hosted UI, API, Self-hosted) is evaluated independently.
- No behaviour observed on one platform or surface SHALL be projected onto another.
- No speculative downgrades are permitted.
- Absence of evidence SHALL be recorded as '?'.
- Later surveys record deltas; they do not invalidate earlier versions.
- Training, fine-tuning, or vendor “memory” features are non-authoritative.
---
III. Evidence Acceptance Rules
Permitted evidence:
- Direct CM artefact testing (MWDUMP, TMLDUMP, XDUMP)
- Author-observed behaviour
- Publicly available platform documentation
Excluded evidence:
- Marketing material
- Roadmaps or intent statements
- Community anecdotes
- Unverifiable benchmarks
---
IV. Search Predicate Invariants
The following search predicates SHALL be reused unchanged in future survey executions unless explicitly amended by the curator.
They are expressed parametrically using {PLATFORM} and {SURFACE}.
Core Capability Predicates
- "{PLATFORM} file upload supported formats limits"
- "{PLATFORM} semantic document ingestion"
- "{PLATFORM} read uploaded file context behaviour"
- "{PLATFORM} deterministic prompt reproducibility"
- "{PLATFORM} context reset new session behaviour"
- "{PLATFORM} system prompt governance control"
Artefact Egress Predicates
- "{PLATFORM} export generated file"
- "{PLATFORM} save model output as file"
- "{PLATFORM} return generated document"
- "{PLATFORM} download assistant output"
- "{PLATFORM} file output api"
Transport and Round-Trip Predicates
- "{PLATFORM} re-upload generated file"
- "{PLATFORM} reuse uploaded document new session"
- "{PLATFORM} file upload reassert context"
- "{PLATFORM} multi-session document persistence"
Surface Differentiation Predicates
- "{PLATFORM} hosted ui vs api differences"
- "{PLATFORM} api file handling"
- "{PLATFORM} batch inference file input"
- "{PLATFORM} cli or sdk document ingestion"
Deployment / Embedded Predicates
- "{PLATFORM} self hosted deployment"
- "{PLATFORM} on premise llm deployment"
- "{PLATFORM} open weights license"
- "{PLATFORM} edge or embedded inference"
- "{PLATFORM} hpc inference deployment"
---
V. Embedded Suitability Clarification (V2+)
On-Site / Embedded Suitability SHALL be evaluated as architectural feasibility of curator-controlled deployment.
It SHALL NOT be interpreted as:
- a measure of CM correctness,
- a requirement for SaaS platforms,
- or a downgrade of CM-capable hosted systems.
A ChatGPT-like system deployed on curator-controlled infrastructure with CM governance SHALL be marked embedded, even if the original SaaS offering is not redeployable.
---
VI. Change Control
Any modification to:
- capability definitions,
- symbol semantics,
- search predicates,
- or execution rules
MUST be recorded as a new curator amendment and versioned separately.
Applicability Note
The invariants and predicates defined in this section SHALL be applied to all market survey executions from Version 2 onward.
They SHALL NOT be used to reinterpret, correct, or invalidate Version 1 results, which remain a valid historical snapshot executed under the definitions and evidence available at that time.
---
I. Capability Definition Invariants (Authoritative)
All capability meanings are fixed by: Cognitive Memoisation: LLM Systems Requirements for Knowledge Round Trip Engineering.
No platform documentation, analyst interpretation, or inferred behaviour SHALL override these definitions.
The following requirements are locked:
- Semantic Artefact Ingress
- Durable Artefact Egress
- Context Determinism
- Transport Independence
- Governance Compatibility
- On-Site / Embedded Suitability (deployment classification only)
Symbol semantics are fixed as:
- ✓ (satisfies requirement)
- ✗ (fails requirement)
- ? (insufficient evidence to determine)
Successful CM round-trip (emit → capture → re-ingest → assert → govern) SHALL be treated as decisive evidence for:
- ingress,
- egress,
- determinism,
- and transport independence.
---
II. Survey Execution Invariants
- Each platform surface (Hosted UI, API, Self-hosted) is evaluated independently.
- No behaviour observed on one platform or surface SHALL be projected onto another.
- No speculative downgrades are permitted.
- Absence of evidence SHALL be recorded as '?'.
- Later surveys record deltas; they do not invalidate earlier versions.
- Training, fine-tuning, or vendor “memory” features are non-authoritative.
---
III. Evidence Acceptance Rules
Permitted evidence:
- Direct CM artefact testing (MWDUMP, TMLDUMP, XDUMP)
- Author-observed behaviour
- Publicly available platform documentation
Excluded evidence:
- Marketing material
- Roadmaps or intent statements
- Community anecdotes
- Unverifiable benchmarks
---
IV. Search Predicate Invariants
The following search predicates SHALL be reused unchanged in future survey executions unless explicitly amended by the curator.
They are expressed parametrically using {PLATFORM} and {SURFACE}.
Core Capability Predicates
- "{PLATFORM} file upload supported formats limits"
- "{PLATFORM} semantic document ingestion"
- "{PLATFORM} read uploaded file context behaviour"
- "{PLATFORM} deterministic prompt reproducibility"
- "{PLATFORM} context reset new session behaviour"
- "{PLATFORM} system prompt governance control"
Artefact Egress Predicates
- "{PLATFORM} export generated file"
- "{PLATFORM} save model output as file"
- "{PLATFORM} return generated document"
- "{PLATFORM} download assistant output"
- "{PLATFORM} file output api"
Transport and Round-Trip Predicates
- "{PLATFORM} re-upload generated file"
- "{PLATFORM} reuse uploaded document new session"
- "{PLATFORM} file upload reassert context"
- "{PLATFORM} multi-session document persistence"
Surface Differentiation Predicates
- "{PLATFORM} hosted ui vs api differences"
- "{PLATFORM} api file handling"
- "{PLATFORM} batch inference file input"
- "{PLATFORM} cli or sdk document ingestion"
Deployment / Embedded Predicates
- "{PLATFORM} self hosted deployment"
- "{PLATFORM} on premise llm deployment"
- "{PLATFORM} open weights license"
- "{PLATFORM} edge or embedded inference"
- "{PLATFORM} hpc inference deployment"
---
V. Embedded Suitability Clarification (V2+)
On-Site / Embedded Suitability SHALL be evaluated as architectural feasibility of curator-controlled deployment.
It SHALL NOT be interpreted as:
- a measure of CM correctness,
- a requirement for SaaS platforms,
- or a downgrade of CM-capable hosted systems.
A ChatGPT-like system deployed on curator-controlled infrastructure with CM governance SHALL be marked embedded, even if the original SaaS offering is not redeployable.
---
VI. Change Control
Any modification to:
- capability definitions,
- symbol semantics,
- search predicates,
- or execution rules
MUST be recorded as a new curator amendment and versioned separately. Absent such amendment, these invariants remain binding.