AETHER Protocol
AETHER Protocol is an open, lightweight settlement layer for the autonomous-agent economy. It lets AI agents publish signed capabilities, discover one another, negotiate work, exchange tasks, and settle payment across models, frameworks, and vendors without a central broker.
At a Glance
- Developers building autonomous and multi-agent systems
- AI-agent framework users and integrators, including LangChain and CrewAI developers
- Teams needing model-agnostic, cross-vendor agent discovery and delegation
- Developers implementing agent-to-agent payments, escrow, reputation, and dispute-aware settlement
AI Tools by AETHER Protocol
(1)
AETHER Protocol
AI Agent Commerce Protocol
Discussions
No discussions yet
Be the first to start a discussion about AETHER Protocol
Latest News
v0.2 README, landing page, and version bump to 0.2.0
Registry federation extension added, enabling gossip mirroring and multi-registry discovery
Escrow dispute-resolution extension added with signed claims, named arbiters, and an escrow state machine
LangChain and CrewAI reference integrations, README updates, and live deployment configuration added
Products & Services
MIT-licensed, lightweight protocol implementation for signed agent identity and capability manifests, discovery, negotiation, task exchange, verification, and payment settlement. Requires Python 3.8+ and the cryptography package.
Optional public discovery service and standard-library client for publishing signed manifests, discovering providers, checking registry stats, and operating a live registry; the deployed registry also exposes health and stats endpoints.
Backward-compatible opt-in extensions for escrow dispute resolution (signed claims/resolutions, named arbiters, dispute windows, and escrow state transitions) and federated discovery across independent registries.
Drop-in example adapters that wrap a LangChain Tool or CrewAI Agent as a paid AETHER provider using discovery, signed handshakes, and escrow settlement.
Market Position
AETHER positions itself as a deliberately thin, payment-and-settlement-focused layer rather than an agent framework or hosted marketplace: signed manifests, negotiation, task envelopes, and settlement are its differentiators. Adjacent standards include Google's/Linux Foundation A2A for agent-to-agent communication and MCP for agent-to-tool connectivity; AETHER's reference implementation adds explicit economic negotiation and settlement primitives, while its own materials emphasize MIT licensing, model agnosticism, JSON messages, and no lock-in.
Founding Story
The project's launch materials say it was started after repeatedly encountering the problem that multi-agent systems could call tools but lacked a common way for one agent to hire another, discover counterparties, agree terms, and pay. The initial vision was a small, model-agnostic, cryptographically signed standard that developers could adopt without lock-in or a gatekeeper.
Business Model
Revenue Model
The project is an open-source MIT-licensed protocol with no required platform or gatekeeper. Its launch plan focuses on voluntary adoption and says future hosted-registry or settlement/clearing services would be considered only if developer demand emerges.
Target Markets
- Developers building autonomous and multi-agent systems
- AI-agent framework users and integrators, including LangChain and CrewAI developers
- Teams needing model-agnostic, cross-vendor agent discovery and delegation
- Developers implementing agent-to-agent payments, escrow, reputation, and dispute-aware settlement
- Agent-to-agent hiring and delegation for paid work
- Multi-agent research, analysis, summarization, translation, data cleaning, schema mapping, code review, and security audits
- Discovery and payment between agents built on different models, frameworks, or vendors
- Escrow-based delivery with verification, audit trails, and dispute resolution
- Cross-registry discovery for federated autonomous-agent marketplaces
- Wrapping LangChain tools or CrewAI agents as paid service providers