How Verification Works

Overview

Nexr verification exists to prove that a builder's work is authentic, active, and contains real technical or creative depth. Unlike traditional self-reported resumes that are easy to inflate or fake, the Nexr Profile programmatically verifies builds directly at the source.

When a build is submitted, our backend initiates multi-layered automated pipelines tailored specifically to the build category—querying repository APIs, extracting media and design file metadata, parsing file structures, and executing multimodal AI analysis.

Code & Repository Engine

For projects imported from a version control platform like GitHub, verification runs a multi-stage check to prove authorship, stack depth, and codebase health:

• OAuth REST Integration: We query GitHub APIs using authenticated tokens to fetch repository details, commit history stream, language distributions, and public contributors list.
• Ownership Confidence: We evaluate commit history ratios, PR activity, and file contributions to assign accurate builder roles (e.g. Primary Builder vs. Contributor).
• Deep Stack & Dependency Detection: Automatically parses package manifests (package.json, requirements.txt, Cargo.toml) to detect frameworks, libraries, and tools (Next.js, TailwindCSS, PyTorch, Docker, etc.).
• Structure & Boilerplate Filtering: Rejects empty repositories, single-file bookmarks, or pure configuration templates. Analyzes folder trees, test suite presence, and documentation depth.
• AI Architectural Review: Gemini evaluates file trees, manifests, and documentation to generate architectural summaries and technical confidence scores.

Design & Creative Engine

For UI/UX design flows, brand identities, and graphic assets, verification parses source files and visual output:

• Source Bundle Extraction: Inspects editable design source files (.fig, .psd, .ai, .sketch, .blend, .png, .jpg) to verify raw asset counts, file formats, and source bundle validity.
• Technical Canvas Metadata: Extracts exact canvas dimensions, color mode (e.g., RGB / CMYK), channel counts, and source file structure.
• Color Palette & Hex Extraction: Programmatically parses image color data to generate dominant color palettes with interactive #HEX copy pills.
• Multimodal AI Vision Critique: Gemini Vision analyzes design preview screens to evaluate visual hierarchy, typography choices, component consistency, and UX flow structure.
• Interactive Gallery & HD Lightbox: Displays preview assets in an interactive gallery equipped with a full-screen Lightbox modal and right-click visual protection.

Video & Motion Engine

For video edits, motion graphics, and 3D animations, verification inspects technical video streams and channel origins:

• Media Stream Parsing: Technical extraction of video resolution (1080p, 4K), framerate (24fps, 30fps, 60fps), duration, and audio/video codecs.
• Channel & Platform Attributing: Fetches creator handles, view metrics, and publish timestamps via oEmbed and platform APIs (YouTube, Vimeo, TikTok) and matches channel identities against verified social profiles.
• Custom Theatre Experience: Renders videos inside a responsive Theatre Player with custom aspect ratio scaling and metadata overlays.

AI Visual Signature & Anti-Plagiarism

To protect creators against copy-paste plagiarism and stolen portfolio assets, Nexr generates a perceptual fingerprint for every creative build:

• High-Dimensional Vector Hash (VECTOR_PRINT): Images and video keyframes are processed through perceptual hashing algorithms that convert visual channels into unique SHA-256 vector matrices.
• Cross-Network Duplicate Detection: If a build with matching visual channels is re-uploaded by another account, the system flags the collision to safeguard original ownership.

Proof Disclaimers & Transparency

Every build detail page includes a clear, category-tailored verification disclaimer explaining how the metrics were derived:

• Design Builds: "Self-reported by builder. All metrics, color palettes, canvas dimensions, and layer structures above are derived from direct design source file analysis."
• Video Builds: "Self-reported by builder. All metrics, resolution, framerate, and stream metadata above are derived from direct media asset analysis."
• Code Builds: "Self-reported by builder. All metrics and structures above are derived from direct codebase analysis."

Reverification & Rate Limits

Builders can request project updates as their work evolves over time:

• Quick Refresh: Instantly syncs live repository metrics such as stars, forks, views, and external links.
• Full Reverify: Re-executes the complete codebase, AI parsing, color extraction, and vector fingerprinting engine. This operation is rate-limited to 3 requests per 10 minutes per builder to prevent quota abuse.