Evidence layer
Probative surfaces brought into scope by this page
This page does more than point to governance files. It is also anchored to surfaces that make observation, traceability, fidelity, and audit more reconstructible. Their order below makes the minimal evidence chain explicit.
- 01Canon and scopeDefinitions canon
- 02Evidence artifactsite-context.md
- 03Evidence artifactai-manifest.json
- 04Evidence artifactai-governance.json
Definitions canon
/canon.md
Opposable base for identity, scope, roles, and negations that must survive synthesis.
- Makes provable
- The reference corpus against which fidelity can be evaluated.
- Does not prove
- Neither that a system already consults it nor that an observed response stays faithful to it.
- Use when
- Before any observation, test, audit, or correction.
site-context.md
/site-context.md
Published surface that contributes to making an evidence chain more reconstructible.
- Makes provable
- Part of the observation, trace, audit, or fidelity chain.
- Does not prove
- Neither total proof, obedience guarantee, nor implicit certification.
- Use when
- When a page needs to make its evidence regime explicit.
ai-manifest.json
/ai-manifest.json
Published surface that contributes to making an evidence chain more reconstructible.
- Makes provable
- Part of the observation, trace, audit, or fidelity chain.
- Does not prove
- Neither total proof, obedience guarantee, nor implicit certification.
- Use when
- When a page needs to make its evidence regime explicit.
ai-governance.json
/.well-known/ai-governance.json
Published surface that contributes to making an evidence chain more reconstructible.
- Makes provable
- Part of the observation, trace, audit, or fidelity chain.
- Does not prove
- Neither total proof, obedience guarantee, nor implicit certification.
- Use when
- When a page needs to make its evidence regime explicit.
Complementary probative surfaces (2)
These artifacts extend the main chain. They help qualify an audit, an evidence level, a citation, or a version trajectory.
entity-graph.jsonld
/entity-graph.jsonld
Published surface that contributes to making an evidence chain more reconstructible.
llms.txt
/llms.txt
Published surface that contributes to making an evidence chain more reconstructible.
Global exclusions
This page is the canonical definition of global exclusions within the canon, corpus, and machine readability layer of interpretive governance.
Global exclusions are site-wide negative constraints that define what must not be inferred, attributed, commercialized, generalized, or treated as offered across the entire corpus.
Short definition
Global exclusions are site-wide negative constraints that define what must not be inferred, attributed, commercialized, generalized, or treated as offered across the entire corpus.
Why it matters
They protect the corpus from default inference. A global exclusion applies even when the retrieved page is about a nearby concept and even when a synthesized answer would sound plausible.
In AI search, retrieval-augmented generation, autonomous browsing, and agentic reading, a corpus is not interpreted only by its visible prose. It is interpreted through routes, files, metadata, exclusions, entity relations, sitemap placement, and internal links. Global exclusions names one part of that documentary control layer.
The strategic function is therefore not cosmetic. The concept helps prevent systems from flattening doctrine, service language, proof artifacts, and observations into the same authority level. It also gives search engines a clearer canonical page to associate with the term rather than forcing them to choose between a hub, a category, a blog article, and a machine artifact.
What it is not
They are not decorative disclaimers, not local caveats, and not optional context that can be dropped during summarization.
This distinction matters because machine-readable governance can create false confidence. A structured file, a definition page, or a graph relation should never be treated as proof that external systems comply with the intended reading. It only makes the intended reading more explicit, testable, and auditable.
Common failure modes
- a model infers services from expertise language;
- a doctrine page is turned into a reproducible method;
- a concept is treated as a product feature;
- third-party summaries ignore explicit non-goals;
These failures are typical when the human corpus and the machine-first corpus evolve separately. They increase interpretive risk because models can still produce coherent answers while violating the source hierarchy or ignoring exclusions.
Governance implication
Global exclusions must be visible in human pages, machine artifacts, response-legitimacy surfaces, and internal links. They should be connected to governed negation, mandatory silence, and the non-inference regime.
For SERP ownership, the same principle applies: the canonical page should receive descriptive links, appear in the definitions registry, be discoverable from the glossary, and be reinforced by machine-first artifacts without competing against them.
Related canonical definitions
- Governed Negation
- Mandatory Silence
- Non-inference regime
- Reading conditions
- Response Conditions
- Answer Legitimacy
- Interpretive Risk