How proofxyz/moonbirds-assets works
Project type and purpose
• This is an open-source “asset repository” for Moonbirds NFT traits and artwork. There is no running frontend or backend service here—this repo simply houses the raw materials (PNGs, SVGs, and trait metadata) that downstream systems (e.g., image-composition pipelines, NFT metadata servers, or static‐site generators) will consume.
File structure
• Top-level files:
• README.md – describes purpose
• traits.json – a central metadata file listing all trait categories and possibly their weights or IDs
• Directories:
• traits/ – subfolders (Beak, Body, Eyes, etc.) each containing the source images for that trait category
• collection/png and collection/svg – the output folders, presumably containing fully composed Moonbird images in PNG and SVG formats
• No build scripts or config files—this is purely an assets layout.
Components
a. Trait Source Store
• Traits JSON (traits.json)
• Trait directories (traits/Beak, traits/Body, …) with PNG/SVG files
b. Asset Generation Pipeline (external or hypothetical)—a service or script not in this repo that:
• Reads traits.json
• Randomly (or deterministically) selects one file per trait category
• Composes them into a final image layer stack
• Outputs into collection/png and collection/svg
c. Collection Store
• Output directories (collection/png, collection/svg)
d. Downstream Consumers (external)
• NFT metadata server
• Static website or marketplace front end
• IPFS pinning service
Important architectural patterns and principles
• Single source of truth for trait definitions (traits.json)
• Clear separation of source assets vs. generated output
• Extensibility: add a new trait simply by dropping images into a new folder under traits/ and updating traits.json
• Format parity: ensure any new trait is available in both PNG and SVG if needed
Final reminder
• Simplicity of this repo: it’s purely a data store; the “logic” lives off-repo in an image-composer or CI/CD job