Skip to content

mindful-computer/feed-protocol-whitepaper

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 

Repository files navigation

Feed Protocol Whitepaper

License: MIT Version

Overview

The Feed Protocol is an open, interoperable standard for curating and distributing information intentionally across the web. It empowers both humans and AI agents to design their own information flow — across social, news, commerce, and on-chain ecosystems — without dependence on opaque algorithms or centralized feeds.

🎯 Vision

Feed envisions a world where information isn't something that happens to you but for you. Where feeds no longer dictate your reality, but reflect your awareness.

The protocol serves as the foundation for this shift — a mindful information layer that empowers users and AI agents to curate the signals that truly matter.

🚀 Key Features

  • Open & Interoperable: Public, composable, and permissionless by design
  • User Sovereignty: Data, filters, and models belong to users, not platforms
  • AI-Native: Built for both human and agentic interaction
  • Universal Addressing: feed:// URI scheme for standardized feed access
  • Privacy-First: Local-first personalization preserving user autonomy
  • Extensible: Developers can build apps, agents, and tools atop a common protocol

📋 Core Principles

  1. Openness — Public, composable, and permissionless by design
  2. User Sovereignty — Data, filters, and models belong to users, not platforms
  3. Interoperability — Seamless across APIs, blockchains, and web standards
  4. Intentionality — Information shaped by human and agentic purpose, not algorithmic impulse
  5. Extensibility — Developers can build apps, agents, and tools atop a common protocol

🏗️ Architecture

Components

  • Feed Nodes: Aggregate and serve data streams to clients and agents
  • Feed Schema: JSON or GraphQL-based format defining FeedItem, FeedSource, FeedIntent, and metadata
  • Auth Layer: OAuth + decentralized identity (DID) for secure, user-centric access
  • Ranking Layer: Optional AI/LLM scoring guided by user or agent intents
  • Storage Layer: IPFS, Arweave, or encrypted cloud delivery to ensure data persistence and portability

Lifecycle

  1. Ingestion — APIs, RSS, blockchain events, or AI crawlers collect data
  2. Normalization — Formats are standardized for unified consumption
  3. Personalization — Local or federated agents filter and curate by intent
  4. Distribution — Feeds are published, subscribed to, or monetized openly

🔗 The feed:// URI Scheme

The feed:// URI scheme provides a universal addressing system for both humans and agents to access, subscribe to, and share feeds in a standardized way.

Example URI Structures

  • feed://user/profile?user=user
  • feed://user/timeline?user=user&privacy=public
  • feed://post/view?id=12345
  • feed://collection/favorites?share=user1,user2

Privacy Parameters

  • privacy=public — accessible to everyone
  • privacy=private — only the creator
  • privacy=friends — limited circle
  • share=user1,user2 — share with specific individuals or groups

🛠️ Technical Specifications

Component Description
FeedItem Structured object (title, source, tags, URL, timestamp)
FeedSource Defines origin: API, blockchain, RSS, or manual
FeedIntent Semantic intent describing user or agent's focus
FeedAgent Local or remote LLM that interprets intent and ranks results

APIs & SDKs

  • /feed — Retrieve or filter personalized data
  • /intent — Define or update user goals
  • /agent — Interface between local intelligence and protocol data
  • /subscription — Manage access, monetization, and sharing

Developer Tools

  • TypeScript & Python SDKs
  • Schema registry + validator
  • Sandbox testing environment

🌐 Ecosystem Integrations

Web2 Integrations

X (API tier), Reddit, Spotify, YouTube, BBC, Fandango, and more.

Web3 Integrations

Farcaster, Lens Protocol, ENS, Ceramic, IPFS, and EAS attestation layers.

AI Integrations

Feed Agents can interoperate with frameworks like AutoGPT, OpenDevin, or custom LLMs via /intent.

🗺️ Roadmap

Phase Milestone Focus
1 MVP Protocol Schema + API + open registry
2 Launch Current Flagship app demonstrating mindfulness in practice
3 SDK Release Developer adoption and open-source expansion
4 Agent Integration Default layer for AI agent personalization
5 DAO Governance Decentralized standards and sustainability

📖 Reference Implementation

Current is the flagship consumer app built on Feed Protocol, demonstrating what mindful information interaction looks like.

Example Use Cases

  • My Day: Weather + calendar + headlines + music in one intentional view
  • Creator Watch: Spotify + YouTube + Patreon + X threads for focused updates
  • Portfolio Pulse: CoinGecko + Etherscan + TradingView + Lens for market clarity
  • Discovery: MySpace for a glimpse into users' interests
  • Feed Marketplace: AppStore + MasterClass for exclusive access to private feeds

🤝 Contributing

We welcome contributions to the Feed Protocol whitepaper and implementation. Please see our contributing guidelines for more information.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

📞 Contact

📚 Resources


Feed is the mindful information layer for the agentic internet.
From noise to nuance, it brings meaning back to the signal.

© 2025 Feed Protocol / Mindful Computer

About

Feed Protocol: The Mindful Information Layer for the Agentic Internet

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published