Infrastructure layer · API & MCP
Lifecycle intelligence as structured infrastructure.
The installed asset graph is queryable. Manufacturers consume it via REST today and via MCP from inside their own AI agents tomorrow — the same lifecycle reconciliation, the same confidence metadata, two interfaces.
Every asset state transition — estimate, install, commissioning, service, warranty, replacement — emitted as a structured event.
Read-only access to the installed footprint of your SKUs across the contractor channel, with serial-level resolution.
Rolling regional demand index, environmental stress correlation, and replacement pull-forward windows.
Channel-attributed install scores, callback rates, commissioning compliance, and lifecycle continuity.
Anomaly stream — severity-graded, regionally clustered, SKU-attributed, confidence-scored.
REST surface
Example endpoints
Illustrative payloads. Schema, auth, rate limits, and SLA are finalized per manufacturer partner.
https://api.fix9.app/v1/manufacturer/assetsscope: assets:readSummary
Installed asset graph filtered to a manufacturer's SKU footprint.
Request
curl https://api.fix9.app/v1/manufacturer/assets \ -H "Authorization: Bearer f9_live_••••••••••"
Response · 200
{
"asset_id": "ast_8f3a1c9e",
"manufacturer": "Carrier",
"model_number": "24ACC636A003",
"serial_number": "1923E54812",
"install_region": "TX-Houston",
"lifecycle_stage": "service",
"install_confidence": 0.94,
"warranty_status": "active",
"replacement_forecast_window": "2031-Q3",
"data_freshness_seconds": 412
}Lifecycle reconciliation
From field event to manufacturer intelligence
Every interface — REST or MCP — reads from the same reconciled lifecycle graph. There is one source of truth.
Reconciliation engine · install + commissioning + service + warranty · serial-level resolution
MCP interface · forward-looking
The same intelligence, exposed to your AI agents
Model Context Protocol turns lifecycle intelligence into a structured tool surface. A manufacturer's internal agent — warranty triage, demand planning, channel risk — can query the asset graph directly, with scoped access and confidence metadata preserved.
MCP server
mcp.fix9.app/manufacturer
transport: streamable http · auth: oauth2 + scoped api key
assets.queryFilter the installed asset graph by region, SKU, lifecycle stage, warranty status.
warranty.signals.subscribeStream anomaly events as they cross severity thresholds.
contractor.quality.lookupResolve a contractor's reconciled quality profile by region.
forecast.replacementsProject 24-month replacement curve for a SKU family within a region.
demand.regional.snapshotPull current regional demand index with environmental stress overlay.
Example tool call
Manufacturer agent → warranty.signals.subscribe
{
"jsonrpc": "2.0",
"id": 42,
"method": "tools/call",
"params": {
"name": "warranty.signals.subscribe",
"arguments": {
"manufacturer": "Carrier",
"severity_min": "high",
"regions": ["FL-Miami", "AZ-Phoenix"]
}
}
}
// → streamed result
{
"signal_id": "wsig_4a2b71",
"sku": "24ACC636A003",
"severity": "critical",
"anomaly_score": 0.94,
"regional_cluster": "FL-Miami",
"confidence": 0.88,
"freshness_seconds": 312
}MCP interface is on the roadmap. Design partners shape the tool surface.
Access layer
Keys, scopes, freshness, confidence
Every response carries provenance. Scope-bound keys, per-resource data freshness, and a confidence score on each reconciled object.
f9_live_••••••••••14f9_live_••••••••••15f9_live_••••••••••15f9_live_••••••••••13Strategic preview · production schema, auth, and SLA finalized with design partners.