AgentOps-AI / AgentOps-AI/agentops

MAXIA: Multi-Agent Marketplace Observability A2A transactions

Ouverte
#1,321 3 commentaires 0 réactions 0 personnes assignées Voir sur GitHub
enhancement
Langage dominant
Python
Étoiles
5.8k
Forks
619
Métriques de merge des PR
Aucune PR mergée en 30 j

Description

### 💡 Feature Description and Proposed Solution

Context: https://maxiaworld.app is a live AI-to-AI marketplace where autonomous agents discover, buy, and sell
services using USDC across 14 blockchains (Solana, Base, Ethereum, etc.). We run 17+ sub-agents orchestrated by a CEO
agent, with on-chain escrow, signed intent envelopes (ed25519), and real crypto transactions.

▎ Feature Request: Native AgentOps support for agent-to-agent (A2A) commerce observability:

▎ 1. Transaction-level spans — Track agent-to-agent service purchases as first-class spans (buyer agent → escrow lock
→ seller execution → confirmation), not just LLM calls
▎ 2. Multi-agent session linking — Link sessions across buyer/seller agents in the same transaction (our CEO
dispatches to 17 sub-agents per decision cycle)
▎ 3. On-chain event correlation — Correlate AgentOps sessions with blockchain tx hashes (Solana/EVM) for audit trails
▎ 4. Intent envelope tracking — Support for signed intent envelopes (AIP Protocol) as observable events with signature
verification status

▎ This would make AgentOps the go-to observability layer for the emerging agent economy, not just single-agent LLM
monitoring.

▎ We'd be happy to be a design partner / early adopter for this feature.

### 🤔 Related Problem

Current agent observability tools track LLM calls and tool usage, but have no concept of agent-to-agent commerce —
when Agent A pays Agent B on-chain for a service, there's no way to trace the full transaction lifecycle across both
agents' sessions. As multi-agent marketplaces grow, this gap becomes critical for debugging, billing reconciliation,
and compliance.

### 🤝 Contribution

- [x] Yes, I'd be happy to submit a pull request with these changes.
- [ ] I need some guidance on how to contribute.
- [ ] I'd prefer the AgentOps team to handle this update.

Guide de contribution

Ouvrir le guide de contribution

Évaluation

Cette issue n'a pas encore été évaluée.

Recevez les nouvelles issues par e-mail

Un résumé court des issues GitHub adaptées aux débutants.