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.
Q-Ledger
/.well-known/q-ledger.json
Public ledger of inferred sessions that makes some observed consultations and sequences visible.
- Makes provable
- That a behavior was observed as weak, dated, contextualized trace evidence.
- Does not prove
- Neither actor identity, system obedience, nor strong proof of activation.
- Use when
- When it is necessary to distinguish descriptive observation from strong attestation.
Q-Metrics
/.well-known/q-metrics.json
Derived layer that makes some variations more comparable from one snapshot to another.
- Makes provable
- That an observed signal can be compared, versioned, and challenged as a descriptive indicator.
- Does not prove
- Neither the truth of a representation, the fidelity of an output, nor real steering on its own.
- Use when
- To compare windows, prioritize an audit, and document a before/after.
What a phantom URL does not prove
A phantom URL is an interesting but fragile signal. It should be used as an audit indicator, not as total proof about the internal operation of a model.
This clarification sets the limits of the concept so that a useful observation does not become an excessive narrative.
1. It does not prove that a LLM directly crawled the site
A non-existent URL may appear for several reasons: tool-using agent, assisted browser, AI-referred user, crawler, SEO tool, monitoring system, copied link, generative answer, or a combination of several layers.
The log proves that a request occurred. It does not always prove which system generated the route.
2. It does not prove intention
A system does not necessarily “want” to find the page. It may generate a probable path, test a route, transmit an invented URL, follow a suggestion, or act on an output produced elsewhere.
The right formulation is not: “AI wanted this page.”
The prudent formulation is: “a plausible URL was requested in a context compatible with documentary projection.”
3. It does not prove human-like understanding
The coherence of a phantom URL does not mean that the system understands the site like a human reader. It may result from a statistical regularity, slug pattern, lexical relation, or structural completion.
The phantom URL concept does not rely on anthropomorphism. It relies on the observation of structured plausibility.
4. It does not prove that the page must be created
Some phantom URLs expose real latent documentary surfaces. Others expose false expectations, duplicates, confusions, or undesirable routes.
Automatically creating every phantom URL can weaken the site:
- conceptual duplication;
- dilution of the canon;
- multiplication of thin pages;
- validation of a wrong expectation;
- loss of documentary hierarchy.
Creation is only one possible decision among several.
5. It does not prove that the site is incomplete
A phantom URL may reveal a gap. It may also reveal that the content already exists, but that its linking, canonical route, or role in the graph is not explicit enough.
In that case, the right answer is not a new page. It may be a link, clarification, redirect, or coherence map.
6. It does not prove a site error
A phantom URL is not necessarily the symptom of an editorial defect. A highly coherent site may generate projections precisely because it has a readable documentary grammar.
The signal is therefore ambivalent: it shows both structural strength and a possible absence of stabilization.
7. It does not justify automatic redirects
Redirecting every phantom URL to the nearest page is dangerous. An abusive redirect may send a false correspondence signal, blur intent, create low-quality chains, or hide errors that should remain visible.
Redirection should be reserved for cases where the intent is clear and the target page truly answers the implicit need.
8. It must be interpreted at cluster level
An isolated URL may be accidental. A recurring cluster is more significant.
The signal becomes more robust when several phantom URLs:
- belong to the same semantic family;
- follow the same path conventions;
- appear across several periods;
- come from similar contexts;
- point toward the same latent surface.
Final rule
A phantom URL is an indicator of documentary expectation. It does not replace evidence, audit, qualification, or editorial decision.
The right posture is: observe, exclude noise, verify non-existence, measure coherence, cluster, then decide.