Web3 metadata. Trace every label.
Give names, categories, and summaries an origin—and preserve the evidence behind the words your users see.
Readable metadata should not outrun its source. Separate the network-specific token identity from descriptive fields and from classifications introduced by rules, reviewers, or an AI process.
What to define
Field-level provenance
Attach an origin, collection time, and review state to important descriptions.
Versioned interpretation
Preserve label and taxonomy changes without rewriting asset identity.
Controlled enrichment
Allow uncertainty and missing evidence instead of manufacturing a category.
Build three distinct layers
Identity explains which asset a record concerns. Description stores what a source said. Interpretation records what the application added. This separation lets a public display stay readable while a reviewer can still identify where each statement came from and what kind of check was performed.
Keep current and historical views
A current profile selects metadata under a documented policy. A point-in-time export preserves which information was available earlier. Neither should silently substitute for the other. Version source observations, taxonomy definitions, and selection rules so the differences remain reproducible.
Treat descriptive text carefully
Render external descriptions as text, validate field lengths and types, and do not execute instructions embedded in metadata. For AI-assisted outputs, preserve evidence references and a transformation version. This site serves its visual assets locally and uses internal reference notes rather than third-party navigation.
Review the internal methodology and protocol reference notes for the source concepts behind these guides.
Give every label a source.
Web3 Token Metadata: An Evidence-First Provenance Workflow
Separate identity, description, and interpretation while keeping metadata changes, classifications, and evidence reviewable.
Read the guideWeb3 Token Metadata: An Evidence-First Provenance Workflow
Separate identity, description, and interpretation while keeping metadata changes, classifications, and evidence reviewable.
The full guide covers the decisions, limitations, and sample acceptance cases behind this data scope. Bring those questions into the brief before agreeing a production integration.
Read the complete guideYour next build starts with better data.
Define your assets, fields, evidence, and delivery needs. Start with a clear brief—not an assumed service package.