Data Model Overview
Central entities and their relationships — for technically interested users
This chapter is for technically interested users who want to understand how the platform's data is structured.
Central Entities
Organization
Top-level entity. Contains projects, team members, and contracts.
Project
Isolated workspace within an organization. Contains agents, policies, traces, and settings. Each project has its own API keys and rate limits.
Agent
Registered AI agent with: unique ID + API key, risk level, responsibility chain (3 persons), capability model (4 categories × n tools), budgets (3 dimensions), status (Draft → Active → Paused → Suspended → Revoked).
Policy
Governance rule with: neural instruction (natural language), enforcement action (BLOCK / APPROVAL / ANONYMIZE / FLAG), attestation level, scope (all agents / specific agents / specific types).
Trace
Single processed request with: prompt + response, decision (PASSED / BLOCKED / MODIFIED / APPROVED / DENIED), triggered policies, metadata (timestamp, latency, ingestion path), blockchain status (integrity hash, Merkle proof, TX hash).
Contract (Enterprise)
Individual contract with: term, base fee, included requests, overage rates, blockchain quota, auto-renew, expiration warnings.
Platform Connection
Connection to external system (Azure, Google, Slack, Teams, ServiceNow, Jira) with: type/category, OAuth tokens (encrypted), sync status.
Discovered Agent
Agent detected via platform connection, not yet governed, with: source platform, external ID, name, model, tools, instructions, import status (Detected / Imported / Ignored).