did:fid - A Credibly Neutral and Censorship-Resistant DID Method for AT Protocol

A proposal to establish did:fid as the third blessed DID method for AT Protocol, leveraging Farcaster's decentralized identity infrastructure.

Vision

We propose did:fid as a censorship-resistant addition to AT Protocol's supported DID methods. Unlike the existing did:web and did:plc methods, did:fid offers credibly neutral decentralization through smart contracts on Optimism, providing users with sovereign control over their identities.

Background: DIDs in AT Protocol

In AT Protocol, Decentralized Identifiers (DIDs) serve as the foundation of user identity. They function as decentralized accounts. DIDs provide:

• Persistent Identity: Users maintain the same identifier across different services
• Account Ownership: DIDs act as the user's decentralized account, controlling access to their social graph and content
• Cryptographic Verification: DIDs contain public keys for authenticating users and verifying signatures
• Service Discovery: DIDs point to where user data is hosted (PDS - Personal Data Server)
• Provider Independence: Users can migrate between providers of these services while keeping their identity

Currently, AT Protocol supports two DID methods (see specification):

• did:plc - A centralized registry currently controlled by Bluesky PBC
• did:web - Based on DNS/HTTPS, they not suitable for individual user DIDs but great for Brands.

Why did:fid?

The Censorship Problem

Both existing DID methods have critical weaknesses:

• did:plc: Centrally controlled by Bluesky PBC, creating a single point for censoring updates (and in some cases retrieval)
• did:web: Dependent on traditional web infrastructure (DNS, certificate authorities) that can be censored or seized

The Farcaster Solution

did:fid builds on Farcaster's proven identity system:

• 1+ Million Users: Farcaster has over 1 million registered decentralized identities
• Smart Contract Based: Runs entirely on Ethereum L2 blockchain, ensuring credible neutrality
• No Central Authority: No single entity can censor or revoke identities
• Battle-Tested: Operating successfully in production for years
• Recovery & Security Balance: Farcaster IDs provide an optimal balance between self-custody and recoverability. Users control their custody address (can transfer ownership), recovery address (can recover if custody key is lost), and signer keys (for day-to-day operations). This three-tier security model prevents both accidental loss and unauthorized access (learn more)

Implementation

This repository demonstrates how Farcaster DIDs can integrate with AT Protocol:

Current Features

• ✅ Resolves existing did:fid identifiers to W3C-compliant DID Documents
• ✅ Retrieves on-chain custody addresses, recovery addresses, and signing keys
• ✅ Integrates seamlessly with AT Protocol's DID resolver pattern
• ✅ CLI tool for testing and exploration

Architecture

Optimism Smart Contracts:
├── ID Registry (0x00000000Fc6c5F01Fc30151999387Bb99A9f489b)
│   └── Manages FID ownership and custody
├── Key Registry (0x00000000Fc1237824fb747aBDE0FF18990E59b7e)
│   └── Tracks signing keys for FIDs
└── [Proposed] AT Proto Registry
    └── Stores AT Proto handles, verification methods, and PDS endpoints

Missing Piece

One additional smart contract needs to be deployed to store AT Protocol-specific data:

• AT Proto Handle: User's handle in the AT Protocol network
• Verification Method: AT Protocol signing keys
• PDS Service Endpoint: Location of user's Personal Data Server

The current implementation includes placeholders for this data.

Economics

Cost Structure

• Registration: ~$1 to register a new did:fid
• Updates: Less than $0.01 per update
• Lookups: Free

Sustainability Model

• Optional fees for AT Protocol registration could fund:
  • Public AT Protocol infrastructure
  • Open-source development
  • Decentralized relay and indexing services

CLI Tool Usage

Prerequisites

• Bun runtime
• Git for submodule management
• Node.js and npm (for building dependencies)

Setup

# Clone with submodules
git clone <repository-url>
cd farcaster-did
git submodule update --init --recursive

# Build dependencies
bun run setup

# Install packages
bun install

Usage

# Show help and available commands
bun run fid help

# Resolve a Farcaster ID
bun run fid lookup <fid>

# Example
bun run fid lookup 1898

Example Output

{
  "id": "did:fid:1898",
  "verificationMethod": [
    {
      "id": "did:fid:1898#farcaster",
      "type": "EcdsaSecp256k1RecoveryMethod2020",
      "controller": "did:fid:1898",
      "blockchainAccountId": "eip155:10:0x3eFbe95EBdE6042147644Bc39CdfcF54B8E4f523"
    }
  ],
  "service": [
    {
      "id": "#atproto_pds",
      "type": "AtprotoPersonalDataServer",
      "serviceEndpoint": "https://bsky.social"
    }
  ],
  "alsoKnownAs": ["at://fid-1898.bsky.social"]
}

Roadmap

Phase 1: Proof of Concept ✅

• DID resolution for existing Farcaster IDs
• Integration with AT Protocol resolver pattern
• CLI tool for testing

Phase 2: AT Protocol Integration (Future)

• Deploy AT Proto registry smart contract
• Add handle and PDS registration methods

Phase 3: User Tools (Future)

• Web interface for did:fid registration
• Farcaster frame/mini-app for existing Farcaster users to add AT Protocol info

Phase 4: Ecosystem (Future)

• SDKs for multiple languages
• Integration with existing AT Protocol clients
• Documentation and developer resources

Contributing

We welcome contributions to make did:fid a reality! Areas where help is needed:

• Farcaster mini app to for existing users to register and AT Protocol account
• Smart contract development for AT Proto registry
• Testing and security audits
• Documentation and specifications

Technical Details

• Blockchain: Optimism (Layer 2 Ethereum)
• Contracts: Immutable, upgradeable through governance
• Gas Costs: Minimal due to L2 optimization

References

• AT Protocol DID Specification
• Farcaster Protocol
• W3C DID Core Specification

License

MIT - Free to use, modify, and distribute

---

Join us in building a truly decentralized social web where users own their identities.