Why Context-Specific Reputation Beats a Universal Score

A universal score treats all interactions equally, making it trivial for attackers to game the system. A high DeFi lending score offers no defense against a spam attack in a governance forum.

• Sybil resistance requires context-specific cost functions.
• Universal scores create a single point of failure for reputation farming.
• Example: A wallet with a high NFT trading score could be a terrible delegate.

Merging reputation from disparate domains (e.g., DeFi, Social, Gaming) into one number destroys informational value. It's the equivalent of averaging your credit score with your chess ELO.

• Lending protocols need debt repayment history, not follower count.
• Governance systems need proven deliberation, not trading volume.
• Data becomes noise, rendering the score useless for specific decisions.

Context-specific reputation is built from atomic, verifiable claims about specific actions. Think Ethereum Attestation Service (EAS) schemas, not a monolithic score.

• Lending Module: Tracks on-chain repayment history and collateralization.
• Governance Module: Measures proposal quality and voting consistency.
• Applications compose only the attestations they need, creating precise trust graphs.

The Problem: New AVSs (e.g., oracles, bridges) need to bootstrap trust in their node operators. The Solution: EigenLayer's slashing framework creates a context-specific, staked reputation. Operators build credibility within each AVS, not a generic score.

• Capital Efficiency: Operators can reuse stake across AVSs, but slashing is isolated per service.
• Market-Driven Security: AVS developers choose operators based on their service-specific performance and collateral.

The Problem: Sybil attacks and low-quality content plague decentralized social graphs and marketplaces. The Solution: OpenRank, a graph-based reputation algorithm that scores entities (wallets, content) based on the quality of their connections within a specific context (e.g., Farcaster, NFT communities).

• Sybil-Resistant: Reputation is non-transferable and based on network structure, not token holdings.
• Context-Isolated: A user's DeFi lending reputation doesn't influence their social credibility.

The Problem: Off-chain oracle nodes and AI agents have no provable history of reliable execution. The Solution: A zkAttestation system that generates verifiable, context-specific reputation proofs for any off-chain computation (e.g., "This node correctly executed 10,000 ML inferences").

• ZK-Proofs: Reputation is cryptographically verifiable, not just an API call.
• Composable Credentials: Agents can build a portfolio of attestations for different task types (data fetching, model inference).

The Problem: A single "Web3 Score" is meaningless. Lending risk, governance quality, and social influence require orthogonal data. The Solution: Modular reputation primitives. Protocols like Nocturne (privacy), Zero-Knowledge KYC providers, and Safe{Wallet} modules demonstrate that reputation must be a composable, context-specific credential.

• Risk Isolation: A governance attack shouldn't nuke your DeFi credit.
• Composability: Protocols can import only the reputation data they need, avoiding bloated universal graphs.

A single score forces protocols to accept risk from unrelated activities. A validator's DeFi borrowing history is irrelevant to its PoS duties, creating systemic fragility.

• Eliminates Contagion Risk: Isolates failures like MEV extraction from lending collateral quality.
• Prevents Sybil Gaming: Attackers cannot bootstrap reputation from low-stakes contexts (e.g., social DAOs) to attack high-value ones (e.g., bridge validation).
• Enables Granular Slashing: Faults are penalized within the specific vertical (e.g., oracle latency) without nuking a participant's entire standing.

Capital and reputation are unbundled. A service provider can allocate stake and build trust precisely where it creates the most value, mirroring real-world professional licensing.

• Dynamic Stake Weighting: A relayer's reputation on Axelar for cross-chain messages directly influences its bonded requirement, not its unrelated Aave credit line.
• Vertical-Specific Leverage: High-trust actors in one domain (e.g., Chainlink oracle nodes) can't artificially inflate their standing in another (e.g., UniswapX fill competition).
• Optimized Returns on Reputation: Participants earn premium fees in niches they've proven expertise in, creating sustainable service marketplaces.

Context-specific reputation is a primitive that protocols like EigenLayer, Hyperliquid, and Across can compose into bespoke security models without consensus overhead.

• Plug-in Attestation Layers: A bridge can import a validator's context-specific score from a rollup's sequencer set, avoiding redundant verification.
• Interoperable Without Homogenization: Systems like LayerZero's Oracle/Relayer sets or Celestia's data availability committees can maintain independent reputation graphs that interoperate via shared frameworks.
• Faster Iteration: New protocols (e.g., a perp DEX) can bootstrap security by composing trust from established, context-relevant primitive providers (e.g., price oracles, keepers).