🧪
Live sandbox, not mocks
Labs hit real AIMarket oracles, Hub/SDK endpoints, WARDEN, alien-monitor APIs, lottery contracts, and the AI-Factory pipeline — or embedded mocks with the same interfaces.
Not slides about LangChain — runnable labs that call real oracles, open payment channels, scan MCP servers with WARDEN, read GAIA/ATLAS pins, federate hubs, and verify hybrid receipts.
Each academy is a self-contained folder: pytest CI, graded exercises, HTML certificates, step-by-step docs, and Colab notebooks in English, Russian, Spanish, French, and Chinese.
Most AI tutorials stop at calling one model. We teach the layer above — discovery, trust, settlement, and verifiable math — using infrastructure you can run yourself.
Labs hit real AIMarket oracles, Hub/SDK endpoints, WARDEN, alien-monitor APIs, lottery contracts, and the AI-Factory pipeline — or embedded mocks with the same interfaces.
Randomness comes with proofs. Optimization ships certificates. Agents return signed receipts. You learn to audit outputs, not just accept them.
Discover capabilities, open USDC channels, invoke metered APIs, and publish paid services other agents can buy — the same patterns as production AIMarket Protocol v2.
Full EN / RU / ES / FR / ZH localization — lab docstrings, step-by-step guides, notebooks, and certificates. Run locally or one-click in Colab.
Same spine across all academies — pick a track, run labs in order, graduate with a certificate.
Concept in the lab docstring + step-by-step guide. No black boxes.
Local Python or Colab — calls the live ecosystem sandbox for that course.
Graded exercises + HTML certificate. pytest keeps every lab green in CI.
Agent builder · security & crypto · physical world · trust & proofs. Deepen the same spine — don't invent ten more random courses.
Hands-on Python course on AI agent orchestration patterns and agent economy.
Hands-on course on AIMarket Protocol v2, SDK, Hub, escrow, channels, and monetizing agent capabilities.
Hands-on course on the AI-Factory pipeline: public status endpoints, shipped products, stage flow, and embedded mock factory.
Hands-on course on AIMarket Hub federation: live well-known & peers probes, open vs closed door, preview quarantine, declared vs read counts, and approve→crawl admission.
Hands-on course on MCP threat models, static scanning, and the WARDEN firewall for agent safety.
Hands-on course on hybrid signatures in the AIMarket stack: why PQ alone is not enough, how dual keys bind a receipt, and how to verify without trusting the hub.
Hands-on course on verifiable randomness, VDFs, ECVRF, and time-lock puzzles using live AIMarket oracles.
Hands-on course on lottery smart-contract concepts: fair draws, Chronos VDF verification, relayer rounds, and escrow channels.
Hands-on course on GAIA live relays and ATLAS operator maps — provenance, allowlists, sensor honesty, and commercial-rail licence gates.
Hands-on course on the alien-monitor ecosystem visualizer — REST probes, graph models, and React Three Fiber scene mapping.
Hands-on course on physics-inspired AIMarket oracles — Murmuration, Colony, Turing, Ablation, and Landauer with live calls.
Hands-on course on graph-based trust: EigenTrust, percolation thresholds, Fiedler connectivity, robust consensus, and sandpile cascades.
Hands-on course on combinatorial and continuous optimization with verifiable quality certificates from live AIMarket oracles.
Hands-on course on verifiable randomness, VDFs, ECVRF, and time-lock puzzles using live AIMarket oracles.
Hands-on course on hybrid signatures in the AIMarket stack: why PQ alone is not enough, how dual keys bind a receipt, and how to verify without trusting the hub.
Clone the monorepo, pick a folder, pip install -e ".[dev]", run pytest -q. The same SDK code works against a local hub and a production AIMarket deployment.