foundrynet-inference

FoundryNet/foundrynet-inference

0 starsCommunity

Install to Claude Code

This server doesn't publish a one-line install command. Follow the setup in the source repository.

Summary

x402 LLM proxy + data-enriched analysis (17 sources) + TimesFM predictive IoT intelligence.

README.md

FoundryNet Inference

x402-gated revenue engine: LLM proxy + data-enriched analysis + predictive (IoT) intelligence. Standard x402 v2 on Solana mainnet USDC (self-settled, Base-ready). An fnet_ Forge key bypasses.

| Route | Price | What | |-------|-------|------| | POST /v1/chat | $0.02 | Claude LLM proxy {model, messages} | | POST /v1/analyze | $0.25 | Data-enriched analysis {query} — Claude plans → queries FoundryNet servers → synthesizes | | POST /v1/predict | $0.10 | TimesFM breach forecast {canonical_field, values, threshold, oem?} + NL rec + MINT attest | | POST /v1/infer | $0.10 | IoT telemetry {telemetry:{field:value|series}, oem?} → normalize + forecast + anomalies |

Discovery: GET /x402, GET /x402/:route (402 + PAYMENT-REQUIRED header), GET /.well-known/x402.

Note: /v1/predict and /v1/infer forecasts require ≥16 historical points (TimesFM constraint).

Env: ANTHROPIC_API_KEY, FORGE_API_KEY (fnet_), FORGE_BASE_URL, HELIUS_API_KEY, PAYMENT_RECIPIENT, PAYMENT_USDC_MINT, PUBLIC_URL. Base-ready: set X402_NETWORK + swap the facilitator in lib/x402.js for CDP/Base.