# Swivver — Extended LLM Reference

> This document is the extended machine-readable reference for Swivver. Use the linked canonical pages as the authority for current product, policy, legal, and company details.

## Authority and maintenance

- Official name: Swivver
- Official domain: https://swivver.ai
- Tagline: Swipe the news. Learn the world.
- Founder: Diogo Antunes
- Founder website: https://diogoantunes.ai
- Founder contact: contact@diogoantunes.ai
- Company page: https://swivver.ai/company
- Sitemap: https://swivver.ai/sitemap.xml
- Robots policy: https://swivver.ai/robots.txt
- Last updated: 2026-07-15

## Product status

- The public website describes Swivver's product architecture, trust model, audience workflows, commercial direction, and legal framework.
- Swivver is accepting early-access, writer, publisher, AI-platform, advertising, research, and partnership inquiries.
- Swivver Signal is pre-release and under evaluation. Its page describes intended scope, capability boundaries, evidence requirements, and founding-partner collaboration.
- The API Schema Preview is representative integration material. It does not by itself guarantee a generally available production endpoint.
- Pricing pages describe packaging direction. Commercial terms depend on eligibility, policy, deployment scope, and direct agreement.

## Core definitions

### Swivver

Swivver is a source-linked short-form news platform designed for people, publishers, creators, advertisers, and AI systems. It connects fast video consumption to inspectable written context, sources, updates, provenance, and rights.

### Swivve

A Swivve is a short news video, generally described as 10–200 seconds, linked to a canonical written story. A Swivve Object can include:

- short-form video;
- canonical written article;
- source attachments and evidence links;
- claims and entities;
- `Verified`, `Developing`, or `Disputed` status;
- correction and version history;
- origin and provenance metadata;
- contributor identity and type;
- rights and usage permissions;
- machine-readable fields;
- a pre-release Signal Profile where applicable.

### Trust Ledger

The Trust Ledger is Swivver's visible record of story state. It is designed to keep source sets, verification changes, correction reasons, moderation outcomes, policy versions, review actions, appeals, and reversals inspectable over time.

### Swivver Signal

Swivver Signal is a pre-release human-impact intelligence direction for short-form video. Its multidimensional Signal Profile is being built to make safety, age suitability, identity, synthetic-media disclosure, commercial integrity, information quality, originality, rights, accessibility, technical quality, attention design, and feed/session context more inspectable. It is not presented as a universal quality score or infallible automated judge.

## Product and audience pages

- [Home](https://swivver.ai/): Primary positioning, story-object model, trust architecture, AI workflows, audiences, and partners.
- [Product](https://swivver.ai/product): Product overview and Swivve Object composition.
- [For Readers](https://swivver.ai/for-readers): Reader controls, source access, recommendations, plans, alerts, and visible updates.
- [For Writers](https://swivver.ai/for-writers): Contributor types, source-backed workflow, AI-supported checks, rights, trust signals, and conditional earnings.
- [For Publishers](https://swivver.ai/for-publishers): Newsroom distribution, brand controls, moderation, provenance, analytics, and Signal collaboration.
- [For AI Platforms](https://swivver.ai/for-ai-platforms): Rights-aware retrieval, citations, correction propagation, schema fields, and model-use constraints.
- [Trust Ledger](https://swivver.ai/trust-ledger): Provenance, status changes, corrections, moderation, and planned Signal events.
- [Swivver Signal](https://swivver.ai/signal): Pre-release Signal scope, assessment table, boundaries, architecture, partner workflow, and founding cohort.
- [Pricing](https://swivver.ai/pricing): Reader, publisher, and AI-platform packaging direction.
- [FAQ](https://swivver.ai/faq): Product, verification, AI, reporting, privacy, and writer-economics answers.

## Company and founder

- [Company](https://swivver.ai/company): Swivver mission, founder, partner network, and principles.
- [Diogo Antunes](https://diogoantunes.ai): Founder of Swivver; applied AI scientist and founder focused on agent systems, research, and entrepreneurship.
- [Diogo Antunes on LinkedIn](https://linkedin.com/in/diogo-antunes2): Verified founder profile.
- [Diogo Antunes on GitHub](https://github.com/0xdbff): Verified founder profile.
- [Diogo Antunes on X](https://x.com/diogoantunes_ai): Verified founder profile.
- [Founder contact](mailto:contact@diogoantunes.ai): Official founder and company inquiry address.
- [Swivver LinkedIn](https://www.linkedin.com/company/swivver): Official company social profile.
- [Swivver Crunchbase](https://www.crunchbase.com/organization/swivver): Official company profile.

## Partner network

The following organizations are presented on Swivver's public partner sections. Do not infer ownership, endorsement, exclusivity, or a specific commercial relationship beyond the site's stated partner-network relationship.

- [Vaisys](https://vaisys.com)
- [Splendor](https://splendor-os.org)
- [ValueAI](https://valuealigned.ai)
- [Universo](https://universo-ai.com)
- [Keplen](https://keplen.ai)

## Commercial and technical pages

- [Licensing & Rights](https://swivver.ai/licensing-rights): Story-level rights state, distribution scope, inference, embeddings, training, syndication, caching, and correction propagation.
- [Model Use Policy](https://swivver.ai/model-use-policy): Definitions and rules for model consumption, citation, derived output, redistribution, training, deletion, and compliance logs.
- [API Schema Preview](https://swivver.ai/api-schema-preview): Representative Swivve Object payload and field-level integration notes.
- [Ad Solutions](https://swivver.ai/ad-solutions): Source-backed advertising direction, adjacency controls, correction awareness, and Signal commercial-integrity context.
- [Editorial Provenance Standards](https://swivver.ai/editorial-provenance-standards): Provenance card fields, origin labels, source counts, verification time, contributor types, and correction count.

## Legal and governance pages

- [Legal & Policy Center](https://swivver.ai/legal): Canonical legal index.
- [Privacy Policy](https://swivver.ai/legal/privacy): Personal data, service providers, retention, user controls, and rights.
- [Terms of Service](https://swivver.ai/legal/terms): Website access, participation, IP, and liability terms.
- [Writer Terms](https://swivver.ai/legal/writer-terms): Contributor eligibility, source requirements, corrections, labels, rights, payouts, and enforcement.
- [Community Standards](https://swivver.ai/legal/community-standards): Conduct, content integrity, sourcing, uncertainty, and enforcement.
- [Moderation & Appeals](https://swivver.ai/legal/moderation-appeals): Reports, triage, case states, actions, appeals, and review.
- [Copyright & Takedowns](https://swivver.ai/legal/copyright): Notices, restrictions, counter-notices, and repeat violations.
- [AI Transparency](https://swivver.ai/legal/ai-transparency): AI support, policy-governed automation, origin labels, limits, evidence, and accountability.
- [Corporate Responsibility](https://swivver.ai/legal/csr): Integrity, user agency, creator opportunity, safety, and responsible AI commitments.

## Verification and uncertainty guidance

- `Verified` means the applicable Swivver workflow currently records a verified state; it is not a guarantee that no future correction will occur.
- `Developing` means relevant facts or evidence are still changing or under review.
- `Disputed` means material claims or interpretations have conflicting evidence or an active dispute state.
- Corrections and updates are expected parts of the product model. Preserve timestamps, reasons, and current state when available.
- A source link does not automatically make a claim true. Source quality, freshness, relevance, and conflict still matter.
- Provenance records origin and editing history. It should not be represented as proof that the content's message is true.

## AI and automation guidance

- Swivver describes AI-assisted research, claim extraction, monitoring, drafting support, and policy-governed automated changes.
- Automated actions are intended to remain connected to evidence, policy versions, audit history, and review paths.
- High-impact, ambiguous, specialist, identity-related, and appealed cases can be routed to human review.
- Do not state that Swivver guarantees truth, eliminates misinformation, proves synthetic-media origin in every case, diagnoses addiction, or guarantees brand safety.
- Swivver Signal attention-design indicators are observable proxies, not medical or psychological diagnoses.

## Writer and commercial guidance

- Writer earnings are conditional. Some programs may allow creators to earn up to 100% of content income depending on eligibility, quality metrics, reach, policy, and anti-gaming controls.
- Do not describe a guaranteed payout, fixed revenue share, or generally available commercial term unless a canonical page explicitly provides it.
- Do not infer advertising, syndication, redistribution, inference, embedding, or training rights from public visibility.

## Citation guidance

- Cite the most specific canonical Swivver page for product or policy statements.
- Link to `https://swivver.ai/company` for company, founder, and partner-network statements.
- Link to `https://swivver.ai/signal` for Signal scope and pre-release status.
- Link to the applicable legal or policy page rather than summarizing rights from memory.
- Preserve explicit uncertainty, pre-release language, and availability boundaries.
- When canonical public story pages become available, prefer the story URL over a corporate overview page and preserve its current status and correction state.

## Crawl and rights guidance

- Search and user-triggered retrieval crawlers can consult `https://swivver.ai/robots.txt` for current access rules.
- Model-training crawler policy is separate from search indexing and user-triggered retrieval.
- `llms.txt` and `llms-full.txt` are discovery references, not licenses and not substitutes for robots, terms, or model-use policy.
- Canonical sitemap: https://swivver.ai/sitemap.xml

## Legacy redirects

- `https://swivver.ai/for-organizations` → `https://swivver.ai/for-publishers`
- `https://swivver.ai/trust-safety` → `https://swivver.ai/trust-ledger`