AI Atlas reviews examples that show AI-native taste, technical range, and shipping judgment. Strong submissions include public artifacts, sanitized screenshots, PRs, demos, traces, dashboards, generated reports, datasets, architecture diagrams, or NDA-safe summaries.
Builder fields. Handle, display name, tagline, short bio, avatar initial, public links, and joined date.
Workflow fields. ID, slug, title, summary, readiness status, function area, capabilities, domains, audiences, author ID, evidence, metrics, visuals, privacy notes, and next assets needed.
Canonical axes. Function areas and capabilities come from the Atlas ontology so showcases can be compared instead of only listed. Sanitized and synthetic evidence can be reviewed when the original client work is private.
Collect supplied URLs and source observations without publishing unverified buyer-facing claims.
Check that claims are visible in the source and that screenshots, traces, logs, and datasets are safe to publish.
Accepted entries move into public data and render as builder and showcase pages.
Do not include secrets, credentials, raw production records, personal marketplace messages, private customer or applicant data, non-public financial positions, or identifying media that has not been intentionally approved.
Use sanitized screenshots, synthetic records, aggregate metrics, redacted traces, and public links wherever possible.