Asset Core Documentation
Asset Core is a deterministic spatial-transactional engine that makes world state authoritative, durable, and auditable. When your system needs to know exactly what exists, where it is, and how it changed, Asset Core provides that ground truth.
What You Get
- No duplication, no drift: Every state change commits atomically or fails completely
- Debug by replay: Reconstruct any past state by replaying the commit log
- Audit-grade provenance: Every change is recorded with who, what, when, and why
Decision Gate (Open Source)
Decision Gate is the companion control plane for evidence-backed gates. It evaluates executable goals, emits auditable decisions, and exports runpacks for offline verification. If you need deterministic gates right now, start there. When you need authoritative world state, Asset Core is the substrate.
See It In Action
Game Inventory: Anti-Dup Trades and Replayable Support
Live-service games fail when item duplication exploits slip through, trades partially complete, or support can’t reconstruct what happened. Asset Core makes every equip, trade, and drop a deterministic transaction with stable idempotency keys—anti-dup by default, replayable for support.
View the Game Inventory scenario
Robotics Cell: Deterministic Debugging for Production
Production robotics fails when sensors and planners disagree about ground truth. Asset Core records every sensor correction, position update, and pick/place as a timestamped commit. When something fails, replay the exact sequence—no archaeology required.
View the Robotics Cell scenario
Core Guarantees
- Determinism: Same inputs always produce same outputs—no hidden non-determinism
- Atomicity: Operations succeed together or fail together, no partial corruption
- Bounded operations: 24 well-defined operations make agent integration safe and auditing exhaustive
- Multi-tenant isolation: Each namespace is an independent world with zero cross-tenant bleed
For the formal model, see Runtime Model.
Choose Your Path
Integrate: Building Applications
Wire Asset Core into an application or service. Focus on HTTP contracts, transaction structure, and SDK patterns.
Operate: Running in Production
Responsible for governance, RBAC, and reliability. Focus on namespace management, access control, and troubleshooting.
Automate: Building Agents
Building automation or AI agent systems. Focus on bounded tool calling, safety, and MCP/OpenAI/Gemini integration.
Documentation Structure
- Getting Started — Overview and First Commit
- Concepts — Runtime Model, Containers, Transactions
- API Reference — HTTP API, Action Reference, OpenAPI
- Scenarios — Foundations, Game Inventory, Robotics Cell
- Operations — RBAC, Namespace Governance
- Agents — Tooling Guide, Agent Reasoning
LLM Briefing Pack
Need a single file you can upload into your LLM? The Asset Core LLM Briefing Pack covers the mental model, mapping logic, guarantees, and fit guidance (experimental; website is canonical).
Upload it, describe your domain, and let the model map Asset Core to your use case. You can also point an agent at the canonical URLs (assetcore.io or assetcore.io/docs), though not every page is guaranteed to be visited—the briefing pack keeps the core guidance in one place.
If the model gets stuck or you want to validate its mapping, reach out to us at support@assetcore.io.
Asset Core LLM Briefing Pack (17.5 KB, English-only)