Practical Applications: Proof of Network

(Part 3 of 3 — same chapter in the PDF; split for the web site.)

• Composable complexity. Game mechanics compose: an agent that masters resource gathering and combat individually must learn to coordinate both under economic constraints. This mirrors recipe composition in the Scrypted Network: individual ingredients are simple; orchestrated workflows are complex. • Economic grounding. Web3 games with token economies create real financial consequences for agent decisions. An agent that trades poorly loses real value—the same incentive alignment the network relies on for CRPC staking and attention auctions. Chibi Clash’s auto-battler mechanics provide a controlled environment where agent strategies can be trained, evaluated, and iterated without the latency and cost of real-world deployment. Trained agents become Scrypted ingredients: reusable, versioned, billable capabilities that can be composed into larger workflows. 10.3.2 EVE Online Frontier: hackathon validation The team’s EVE Online Frontier hackathon win demonstrated this thesis in a high-profile competitive setting: a decentralized, agent-driven mission system with generative art and interactive NPC dialogue, built on the Scrypted orchestration stack. The win validated that game AI trained in persistent virtual economies produces agents with transferable capabilities— and that the Scrypted orchestration engine handles the real-time coordination these agents require. 10.3.3 Integration opportunity: OpenClaw and agentic frameworks OpenClaw—the fastest-growing open-source AI agent framework (231,000+ GitHub stars by March 2026)—represents a natural integration target. OpenClaw’s local-first Node.js runtime, multi-channel messaging, MCP tool integration, and sub-agent orchestration align closely with the Scrypted Network’s ingress and composition model (§11.5). The integration path: • OpenClaw agents consume Scrypted recipes as MCP tools, gaining access to the full ingredient catalog without re-implementing provider integrations. • Scrypted registers OpenClaw agent instances as ingredients, enabling game-trained agents to participate in the broader network’s attention auctions. • CRPC verification applies to OpenClaw-mediated work the same way it applies to any other ingredient—the verification layer is protocol-agnostic. Similar integration paths apply to ElizaOS, DayDreams, and other agentic frameworks already identified as integration targets (Chapter 14, §14.6). The pattern is consistent: Scrypted provides orchestration, billing, and verification; external frameworks provide agent runtime and domain-specific capabilities. 10.4 From applications to network Together, the three applications exercise complementary layers of the Scrypted Network: 56

Network layer Delula Sidelines Chibi Clash Recipe orchestration ✓ ✓ ✓ Provider abstraction ✓ Fault tolerance ✓ Pre-computation ✓ Agentic composition (MCP) ✓ ✓ CRPC verification ✓ LoRA training pipeline ✓ On-chain reputation ✓ Crypto-economic staking ✓ Attention auctions ✓ World-model training ✓ Agent-as-ingredient ✓ ✓ Framework integration ✓ No single application proves the full thesis. Together, they cover the orchestration engine, the verification protocol, the economic model, the reputation system, and the world-model training pipeline. The progression from Delula (centralized, trust-based, credit-billed) to Sidelines (CRPCverified, staked, oracle-grounded) to Chibi Clash (game-trained agents as network participants) traces the same path the network itself is taking: from a working system to a verifiable one to an intelligent one. Every infrastructure project needs applications that break it. Delula broke the orchestration layer and forced it to become production-grade. Sidelines is designed to break the verification layer and force it to become cryptographically honest. Chibi Clash is designed to break the intelligence layer and prove that game-trained agents produce transferable, composable capabilities. The network benefits from all three. 57

Source: transcribed from the compiled Scrypted Network Design whitepaper PDF for web reading. Layout, figures, and pagination may differ from the PDF.