MechaReg — Fairness Policy

Version: v1.0 Last updated: 21 Apr 2026

Our commitment

MechaReg has no preferred placement. No sponsored rows. No private ranking overrides.

Visibility in MechaReg is determined exclusively by the quality of a Publisher's data — structure, specificity, evidence, recency, cryptographic signatures, and coherence. No business can buy placement above another. Not from us. Not from anyone.

This is not a marketing claim. It is a binding policy, enforced at the code level, with public disclosure on violation.

1. What this Policy covers

This Policy applies to every surface where MechaReg Profiles are presented or made discoverable:

• The MechaReg homepage directory (mechareg.com/)
• The /sitemap.xml and /ai.json profile indices
• The /llms.txt and /llms-full.txt registry guides
• Any public API endpoint that returns Profile lists or search results
• Any Mecharim-operated Mecha that recommends or surfaces Profiles to

external AI systems or human users

It covers all Publisher organizations equally — regardless of Plan, spend, tenure, size, geography, or industry.

2. The visibility signals

Profile visibility, ordering, and recommendation within MechaReg surfaces are determined by the following signals — and only these signals:

2.1 Structure (high weight)

How completely the Publisher has populated the canonical profile model. A profile with all structural fields filled (legal name, short description, address, contacts, registry tree with Bases, Units, Anchors, and Xenkeys, Mecha addresses) scores higher than a sparse profile.

2.2 Specificity (high weight)

How precise and falsifiable the published statements are. A Xenkey that says "cold-rolled 304 stainless, 0.5–3.0 mm thickness, SGS-certified, MOQ 500 kg, FOB Shanghai, 14-day lead time to EU" scores higher than one that says "we sell steel."

2.3 Evidence (high weight)

How well the published claims are supported by verifiable evidence. Xenkeys that include evidence links — certificates, inspection reports, data sheets, regulatory filings — with content hashes enabling integrity verification, score higher than unsupported claims.

2.4 Recency (moderate weight)

How recently the Profile was refreshed from the Mecharim platform. Profiles exported within the last 7 days score highest. Profiles older than 30 days carry a staleness marker and are deprioritized.

2.5 Cryptographic signatures (moderate weight)

Whether the Profile carries a valid cryptographic signature from the Publisher's Crew, binding the content to a verified Origin. Signed Profiles score higher than unsigned ones. This signal also gates the trust level: a valid signature is required for the "verified" trust tier.

2.6 Coherence (moderate weight)

Whether the Profile is internally consistent. A Profile whose Xenkeys contradict each other, whose address does not match its claimed country, or whose Mecha addresses reference nonexistent Crews scores lower than a coherent one. Coherence is computed automatically from structural constraints in the data model.

2.7 Completeness of the registry tree (low–moderate weight)

Whether the Publisher has built out its hierarchy beyond the minimum. A Profile with multiple Bases, Units with typed Anchors, and Xenkeys describing each Anchor scores higher than one with only the minimal required structure.

3. What is explicitly excluded

The following factors are never used to determine Profile visibility, ordering, or recommendation:

4. External AI consumption

MechaReg Profiles are consumed by external AI systems — ChatGPT, Claude, Gemini, Perplexity, DeepSeek, and millions of private procurement and search agents. Mecharim does not control how external AIs rank, cite, or recommend Profiles. Our commitment covers only Mecharim-operated surfaces.

However, we design MechaReg to make quality legible to external systems: every Profile carries its trust level, staleness status, provenance chain, and structured signals in machine-readable form. An external AI that chooses to prioritize quality will find the signals it needs, in the formats it expects, without ambiguity.

5. Mecharim's own promotion Mechas

Mecharim operates a set of public-facing Mechas whose function is to help buyers and AI agents find relevant Publishers across the Registry. These Mechas:

• Use only the signals listed in Section 2 to match and recommend
• Do not favour any Publisher based on the excluded factors in Section 3
• Are subject to the same code-level enforcement as every other Registry

surface

6. Enforcement and audit

6.1 Code-level enforcement

The ranking and recommendation logic is implemented in code that can be audited. There is no admin panel that overrides visibility. No feature flag that promotes a specific Publisher. No database column for "paid placement."

6.2 Public disclosure

If Mecharim ever violates this Policy — through error, misconduct, or external pressure — we will publicly disclose:

• What happened
• Which Publishers were affected
• How long the violation persisted
• What corrective measures were taken
• What prevents recurrence

We make this commitment in advance, in public, in a binding Policy, precisely because we want to be held to it.

6.3 Reporting suspected violations

If you believe a Mecharim-operated surface is ranking, recommending, or displaying Profiles in a way inconsistent with this Policy, contact us at fairness@mecharim.com.

We will investigate and respond within five business days. If a violation is confirmed, we will apply the disclosure procedure in Section 6.2.

7. Changes to this Policy

This Policy can only be changed to add signals or strengthen protections. The list of excluded factors in Section 3 can only grow — it will never shrink. If we add a new signal to Section 2, it must be objectively measurable from Profile data and equally applicable to all Publishers.

Any change to this Policy will be:

• Published on this page with an updated version and date
• Notified on the MechaReg homepage for at least 30 days
• Logged in a public changelog linked from this page

8. Summary

MechaReg is operated by Mecharim Ltd. Infrastructure for the machine-intelligence economy.