Open protocol prototype for AI agents to quote, authorize, execute, verify, settle, and dispute service transactions. Not a payment rail — the semantic layer around agent-to-agent work.
Agent communication frameworks (MCP, A2A) let agents discover and invoke each other. Payment rails move value. But there's nothing in between: no standard for quoting, authorizing spend, verifying results, generating receipts, or resolving disputes.
Every team inventing their own clearing logic creates fragmentation, incompatibility, and trust failures at scale.
NeuralClear defines standardized objects and HTTP endpoints for the full transaction lifecycle — discovery, quoting, mandate authorization, task execution, proof verification, settlement receipt, and dispute resolution.
It doesn’t issue value, custody funds, or replace payment rails. It defines the commercial semantics and clearing records that make agent transactions verifiable and composable.
Every agent service transaction follows the same lifecycle. Click any stage to explore.
The AI agent stack has emerging standards at every layer. NeuralClear occupies the clearing layer between communication and settlement.
All JSON-serializable, version-tagged, composable across any agent framework.
Full clearing stack locally. HTTP server, SQLite ledger, CLI client, web dashboard.
Clone and install Python dependencies.
pip install -e ".[http]"FastAPI + SQLite on port 8000.
uvicorn server.app:app --port 8000python3 -m client.cli quote request \ agent.pdf_summarizer summarize.pdf
python3 -m client.cli task submit quote_xxx \
--text "Summarize this PDF."python3 -m client.cli balances
Three agents pre-registered in the local sandbox for protocol testing.
Open protocol. Free local stack. Hosted infrastructure coming.
Nine typed objects covering the full agent service transaction lifecycle.
{
"quote_id": "quote_a3f9e2b1",
"spec_version": "0.3",
"provider_agent_id": "agent.pdf_summarizer",
"requester_agent_id": "agent.orchestrator_01",
"capability_id": "summarize_pdf",
"price": 0.05,
"currency": "USD",
"input_hash": "sha256:a1b2c3d4...",
"expires_at": "2024-01-15T12:05:00Z",
"status": "pending_acceptance"
}Get NeuralClear running locally in under five minutes.
# Clone the repository git clone https://github.com/Zihuatanejo63/neuralclear cd neuralclear pip install -e ".[http]"
uvicorn server.app:app --port 8000 # Dashboard at http://localhost:8000/dashboard
python3 -m client.cli agents list
python3 -m client.cli quote request agent.pdf_summarizer summarize.pdf
python3 -m client.cli task submit quote_xxx \
--text "Summarize this PDF text."python3 -m client.cli balances python3 -m client.cli receipts list open http://localhost:8000/dashboard
Full NeuralClear environment locally. HTTP server, SQLite ledger, CLI, web dashboard.
pip install -e ".[http]"uvicorn server.app:app --port 8000python3 -m client.cli agents list
python3 -m client.cli quote request \ agent.pdf_summarizer summarize.pdf
python3 -m client.cli task submit quote_xxx \
--text "Summarize this PDF."python3 -m client.cli balances
Three reference agents pre-registered for protocol testing.
{
"agent_id": "agent.pdf_summarizer",
"spec_version": "0.3",
"name": "PDF Summarizer Agent",
"endpoint": "http://localhost:8001",
"capabilities": [{
"capability_id": "summarize_pdf",
"pricing": { "model": "per_page", "unit_price": 0.005, "currency": "USD" }
}]
}Developed in public as an open protocol prototype. Each milestone adds a production-relevant layer.
Open protocol. Free local stack. Hosted infrastructure coming.