QYN VERIFY — TECHNICAL DOCUMENTATION
QYN Documentation
Open verification infrastructure. Built to be integrated. Technical and operational documentation for deploying QYN as a content verification layer in universities, platforms, and applications.
Overview
What QYN Verify provides
QYN Verify is a protocol-level system for evaluating the integrity of digital content. It accepts arbitrary inputs – URLs, raw text, and file hashes – and produces structured verification records containing trust scores, AI-generation assessments, manipulation signals, and provenance metadata. Each verification result is committed to the QYN blockchain and becomes part of a permanent, auditable ledger.
The protocol is intentionally narrow in scope. It does not attempt to make political, legal, or ethical judgements about content. Instead, it focuses on reproducible signals: source reputation, internal consistency, linguistic properties, cross-referenceability, and tamper evidence. The emphasis is on creating a durable infrastructure primitive that other systems – newsrooms, universities, regulators, and application developers – can build on.
Infrastructure layer
Exposed as a chain-level service rather than an application. Any system capable of signing and submitting transactions can use it.
AI-native
The verification engine is implemented as a native Rust library inside the node, rather than as a call-out to an external AI provider.
Immutable records
Verification outputs are written to the QYN blockchain. There is no administrative mechanism for retroactive modification or deletion.
The Problem
Lack of a native trust layer
The internet stack provides addressing (IP), routing (BGP), naming (DNS), transport security (TLS), and content distribution (CDNs). It does not provide a native notion of whether a piece of content is credible, authentic, or human-generated. All attempts to retrofit such a notion at the application layer have inherited the limitations of centralisation and scale.
Human review
Editorial and fact-checking teams cannot approach the volume and velocity of global information flows. Review decisions are opaque and non-reproducible.
Platform moderation
Large platforms can remove or downrank content, but their policies are centrally defined, jurisdiction-specific, and subject to political and commercial incentives.
Standalone AI detectors
Point-in-time AI detection tools provide binary answers without an accompanying, immutable audit trail. Their outputs are not cryptographically verifiable.
Provenance-only approaches
Metadata-based solutions can establish origin but cannot meaningfully assess factual accuracy, manipulation, or AI-generation likelihood.
QYN treats verification as a first-class network primitive. Verification requests and responses are handled by protocol nodes and persisted on-chain, in the same way that transactions are handled and persisted by a payment network.
Architecture
High-level system layout
QYN consists of three primary components: the blockchain (consensus and data layer), the verification engine (analysis layer), and the public RPC interface (integration layer). All three are implemented in Rust and are designed to run in a single node process for minimal latency and reduced complexity.
quyn-core (blockchain)
- • EVM-compatible execution environment.
- • Block time 3 seconds on testnet.
- • Stores verification records as on-chain data structures.
quyn-intelligence (engine)
- • Stateless, deterministic content analyser.
- • Provides both general and academic-specific analysis routines.
- • Tuned for low-latency operation inside the node.
quyn-rpc (interface)
- • JSON-RPC 2.0 over HTTP/HTTPS.
- • Methods for content verification, recent-history access, and academic analysis.
- • Proxied via nginx and TLS-terminated at rpc.getquyn.com.
Verification Engine
Content analysis model
The verification engine operates on an abstract representation of content, identified by a content type and payload. For URLs, the engine evaluates the domain, historical credibility signals, and structural properties of the page. For text, it evaluates linguistic features, citation structure, and internal consistency. For images and other media, the current testnet implementation focuses on source and metadata; full pixel-level analysis is a roadmap item.
Output fields (general verification)
- trustScore (0–100) — composite score derived from source reputation, content structure, and detected anomalies.
- aiGenerated — categorical value in {HUMAN, AI_GENERATED, LIKELY_AI_GENERATED, UNKNOWN}.
- sourceCredibility — {HIGH, MEDIUM, LOW, UNKNOWN} based on origin and historical behaviour of the source.
- manipulationDetected — boolean flag for patterns associated with synthetic or doctored content.
- verificationId — unique on-chain identifier allowing later retrieval and independent verification.
QYN Academic
Academic integrity analysis
QYN Academic extends the verification engine with routines tailored to academic writing. It is designed for integration into university workflows for essay submission, thesis examination, and research integrity checks.
- • AI probability estimation based on phraseology, sentence uniformity, and vocabulary distribution.
- • Detection of boilerplate phrases commonly produced by large language models.
- • Extraction and heuristic validation of citations, including hallucinated source detection.
- • Plagiarism fingerprinting via document hashes suitable for on-chain anchoring.
Blockchain Layer
Persistence and retrieval
Every verification performed by the protocol is associated with a unique verification identifier and stored on-chain. The storage format is optimised for retrieval and audit rather than for arbitrary querying; analytic workloads are expected to be performed using off-chain indexers.
- • Chain ID: 7779 (QYN Testnet).
- • Consensus: proof-of-stake (testnet configuration).
- • Block interval: 3 seconds.
- • Verification data is addressable via the `verificationId` and can be inspected via the explorer.
API Reference (JSON-RPC)
Core methods
qyn_verifyContent
Submits content for verification and returns a structured verification result.
POST https://rpc.getquyn.com
{
"jsonrpc": "2.0",
"method": "qyn_verifyContent",
"params": [
"https://example.com/article",
"article"
],
"id": 1
}qyn_getRecentVerifications
Returns a bounded list of recent verification records held by the node.
{
"jsonrpc": "2.0",
"method": "qyn_getRecentVerifications",
"params": [],
"id": 1
}qyn_analyseAcademicDocument
Runs the academic analysis pipeline on a text representation of a document and returns an academic report.
Security Model
Threat model and assumptions
The security objective is to ensure that once a verification result has been produced and committed, it cannot be altered, repudiated, or selectively withdrawn. The protocol assumes that a majority of stake on the network is controlled by honest participants in the standard proof-of-stake model.
- • Verification records inherit the security properties of the underlying consensus.
- • There is no administrative backdoor for editing or deleting chain data.
- • Client libraries should verify TLS for RPC connections and pin hostnames where appropriate.
- • For high-assurance deployments, institutions are encouraged to run their own QYN nodes.
Reference Use Cases
Typical integrations
Universities
Integrate QYN Academic into submission portals to generate a permanent integrity record for each essay, dissertation, or thesis. Store only the hash and verification identifier on-chain; underlying documents remain in institutional storage.
News organisations
Register each published article with QYN Verify and expose verification IDs in article templates. Third parties can independently confirm that a circulating screenshot or copy matches the original.
Regulators and auditors
Use verification records as part of supervisory workflows to demonstrate that institutions followed documented verification procedures at specific points in time.
Roadmap
Planned work (non-binding)
- • Mainnet launch with independent security review of core protocol and engine.
- • Expanded media analysis for images and video, including robustness to adversarial inputs.
- • Formal specification of the verification pipeline for independent implementation and review.
- • Public dataset releases to support academic research into verification methods.
Governance
Change management
Parameters that materially affect verification behaviour – such as scoring thresholds and analysis heuristics – are versioned and documented. Changes in these parameters must be rolled out via protocol upgrades and are intended to be reviewable by external researchers and institutional partners.
Compliance and Operational Notes
Deployment considerations
- • QYN Verify is an informational system. It does not provide investment advice or financial services.
- • Institutions remain responsible for legal and disciplinary decisions made on top of verification outputs.
- • For regulated environments, running dedicated nodes and maintaining internal logs is recommended.