SE

scottcjn/elyan-compute-skill

Developer tools
71 stars Quality 54 Trend 18MIT

Conway Automaton skill for accessing Elyan Labs GPU compute marketplace via x402 USDC micropayments

Overview

Conway Automaton skill for accessing GPU compute via x402 USDC micropayments. Part of the Beacon Protocol ecosystem — uses beacon_skill.compute_marketplace and openclaw-x402 for payment. All endpoints return HTTP 402 with x402 payment requirements. Compatible with any x402 client library. Payments in USDC on Base (chain ID 8453). Dual economy: also accepts RustChain RTC tokens (1 RTC = $0.10 USD) via X-RTC-Payment header. 12 GPUs totaling 192GB VRAM across V100 32GB, V100 16GB, RTX 5070, RTX 4070, RTX 3060, and M40 cards. Plus IBM POWER8 S824 with 128 threads and 512GB RAM for large model inference. Hailo-8 TPU and 2x Alveo U30 FPGA for specialized workloads. All infrastructure located in Baton Rouge, LA.

README

Elyan Labs Compute Skill

Conway Automaton skill for accessing GPU compute via x402 USDC micropayments.

Part of the Beacon Protocol ecosystem — uses beacon_skill.compute_marketplace and openclaw-x402 for payment.

Install

Conway agents can install this skill:

installSkill("https://github.com/Scottcjn/elyan-compute-skill", "elyan-compute")

Or fetch the SKILL.md directly:

installSkillFromUrl("https://raw.githubusercontent.com/Scottcjn/elyan-compute-skill/main/SKILL.md", "elyan-compute")

Endpoints

Endpoint Price Hardware
/api/compute/inference $0.01 POWER8 512GB (Ollama)
/api/compute/vision $0.05 BAGEL-7B on V100
/api/compute/tts $0.02 XTTS on RTX 4070
/api/compute/video $0.50 ComfyUI on V100 32GB

Payment

All endpoints return HTTP 402 with x402 payment requirements. Compatible with any x402 client library. Payments in USDC on Base (chain ID 8453).

Dual economy: also accepts RustChain RTC tokens (1 RTC = $0.10 USD) via X-RTC-Payment header.

Hardware

12 GPUs totaling 192GB VRAM across V100 32GB, V100 16GB, RTX 5070, RTX 4070, RTX 3060, and M40 cards. Plus IBM POWER8 S824 with 128 threads and 512GB RAM for large model inference. Hailo-8 TPU and 2x Alveo U30 FPGA for specialized workloads.

All infrastructure located in Baton Rouge, LA.

Ecosystem

License

MIT

View the full README on GitHub

Recommended Tools

Try a different keyword or remove a filter.

Install

npx skillfish add scottcjn/elyan-compute-skill
New to MCP? Start with the mcpskill.org guide →