Agent-First Specification (2026.09)
JSON-RPC 2.0 / MCP
Model Context Protocol (MCP) Sourcing Endpoints
In 2026, enterprise procurement agents powered by DeepSeek, Claude, or ChatGPT autonomously parse supply networks. Nexus-V exposes zero-overhead machine-readable tools for continuous capacity negotiation.
Public Crawler Manifest: /llms.txt
Raw llms.txt
Every route, verified factory record, and technical specification is cataloged within our root llms.txt file following the standardized 2026 generative engine indexing specification.
1. MCP Tool: query_verified_capacity
Allows an AI Agent to find unreserved spindle capacity matching specific machine brands and tolerance grades.
{
"name": "query_verified_capacity",
"description": "Locate audited machine shops with verified unreserved machine spindle capacity for a specified manufacturing process and tolerance threshold.",
"parameters": {
"type": "object",
"properties": {
"process": {"type": "string", "enum": ["cnc_5axis", "injection_cleanroom", "sheetmetal_laser"]},
"material": {"type": "string", "example": "PEEK_ASTM_F2026"},
"min_tolerance_mm": {"type": "number", "example": 0.005},
"target_month": {"type": "string", "example": "2026-10"}
},
"required": ["process", "material", "min_tolerance_mm"]
}
}
2. Direct CLI / Agent Execution Example
curl -X POST https://nexus-v-sourcing.pages.dev/api/v1/capacity \
-H "Authorization: Bearer agent_token_2026" \
-H "Content-Type: application/json" \
-d '{
"process": "cnc_5axis",
"material": "PEEK_ASTM_F2026",
"min_tolerance_mm": 0.005,
"target_month": "2026-10"
}'