Hi — I'm working on OM World, a protocol for a decentralized intent economy. Pneuma's framing — sovereign AI agent protocol layer with ERC-721 identity + x402 micropayments + anti-wash reputation + P2P court — overlaps with all four of our primitives (Intent Schema, Agent Mandate, Execution Proof, Tool Registry) more than any other repo we've seen. Three design questions.
1. ERC-721 vs ERC-8004 for agent identity
ERC-721 gives transferable token-bound identity; ERC-8004 is purpose-built for agents with reputation and KYC hooks. Why 721? Transferability is a feature (an agent's identity can change hands) but also a risk (reputation portability becomes a vehicle for laundering). Did you consciously choose transferability, or was it ergonomic — 721 has the wider tooling base?
2. Anti-wash reputation mechanism
"反洗白" — anti-wash — is the hard part of any agent reputation system. Self-trades and sybil attacks are how on-chain reputation gets gamed. What's Pneuma's detection layer — graph analysis of payment paths, stake-bound attestations from independent parties, time-weighted decay, or a combination? This is the question we're avoiding in our spec by leaving reputation out of v0, but a real Genesis-era protocol can't dodge it forever.
3. P2P court enforcement model
"P2P 法庭" (P2P court) is the interesting term — most decentralized protocols use the word "arbitration" or "council" instead. What's distinct about your model: are the judges peers selected at random from a stake pool (lottery + slashing), nominated by reputation, or rotated by some governance schedule? And what's the enforcement — does the court control on-chain settlement directly, or does it issue rulings that other contracts honor by convention?
The four-primitive overlap is striking. Happy to share where OM World landed and would value the comparison — especially on the anti-wash design which we don't have yet.
Hi — I'm working on OM World, a protocol for a decentralized intent economy. Pneuma's framing — sovereign AI agent protocol layer with ERC-721 identity + x402 micropayments + anti-wash reputation + P2P court — overlaps with all four of our primitives (Intent Schema, Agent Mandate, Execution Proof, Tool Registry) more than any other repo we've seen. Three design questions.
1. ERC-721 vs ERC-8004 for agent identity
ERC-721 gives transferable token-bound identity; ERC-8004 is purpose-built for agents with reputation and KYC hooks. Why 721? Transferability is a feature (an agent's identity can change hands) but also a risk (reputation portability becomes a vehicle for laundering). Did you consciously choose transferability, or was it ergonomic — 721 has the wider tooling base?
2. Anti-wash reputation mechanism
"反洗白" — anti-wash — is the hard part of any agent reputation system. Self-trades and sybil attacks are how on-chain reputation gets gamed. What's Pneuma's detection layer — graph analysis of payment paths, stake-bound attestations from independent parties, time-weighted decay, or a combination? This is the question we're avoiding in our spec by leaving reputation out of v0, but a real Genesis-era protocol can't dodge it forever.
3. P2P court enforcement model
"P2P 法庭" (P2P court) is the interesting term — most decentralized protocols use the word "arbitration" or "council" instead. What's distinct about your model: are the judges peers selected at random from a stake pool (lottery + slashing), nominated by reputation, or rotated by some governance schedule? And what's the enforcement — does the court control on-chain settlement directly, or does it issue rulings that other contracts honor by convention?
The four-primitive overlap is striking. Happy to share where OM World landed and would value the comparison — especially on the anti-wash design which we don't have yet.