a11oy / web /src /data /blueprint.ts
betterwithage's picture
sync(space): complete build context — fix BUILD_ERROR (CTO)
518343a verified
Raw
History Blame Contribute Delete
12.4 kB
import { DOMAIN_PROFILES } from '@szl-holdings/domain-profiles';
export const SEVEN_PRINCIPLES = [
{
step: '01',
label: 'Ingest',
glyph: '↓',
desc: 'Ingest data from multiple systems — market feeds, IoT telemetry, document streams, API webhooks — and unify them into a business ontology.',
},
{
step: '02',
label: 'Understand',
glyph: '◉',
desc: 'Interpret signals and goals using domain intelligence and context memory. Build causal graphs that answer why, not just what.',
},
{
step: '03',
label: 'Plan',
glyph: '⬡',
desc: 'A planner agent decomposes objectives into steps and assigns them to specialized operators — research, execution, risk, and verification.',
},
{
step: '04',
label: 'Act',
glyph: '→',
desc: 'Execution agents call tools and services via the tool fabric. Every call requires proper permissions. Sensitive actions require human approval.',
},
{
step: '05',
label: 'Verify',
glyph: '✓',
desc: 'Verifier agents ensure tasks executed successfully and match expected business outcomes before the proof chain is closed.',
},
{
step: '06',
label: 'Audit',
glyph: '◆',
desc: 'The Proof Chain logs each step, tool call, approval, and result in an immutable, queryable record built for compliance.',
},
{
step: '07',
label: 'Learn',
glyph: '↻',
desc: 'Outcome memory captures what worked and what did not — feeding calibrated confidence back to the next cycle.',
},
] as const;
export type PrincipleLabel = typeof SEVEN_PRINCIPLES[number]['label'];
export const PRINCIPLE_LABELS: PrincipleLabel[] = SEVEN_PRINCIPLES.map(p => p.label);
export const BLUEPRINT_COMPONENTS = [
{
id: 'agent-runtime',
name: 'Agent Runtime',
purpose: "Manages every agent's identity, state, memory, reasoning trace, task lifecycle, and audit logs — the central orchestration engine.",
principles: ['Plan', 'Act', 'Audit'] as PrincipleLabel[],
adjacent: ['Agent Roles', 'Orchestration', 'Agent Identity'],
category: 'Core' as const,
},
{
id: 'business-ontology',
name: 'Business Ontology',
purpose: 'A domain-specific model of objects, events, roles, policies, and outcomes that provides meaning to data and actions across verticals.',
principles: ['Ingest', 'Understand'] as PrincipleLabel[],
adjacent: ['Tool Fabric', 'Memory', 'Agent Runtime'],
category: 'Intelligence' as const,
},
{
id: 'tool-fabric',
name: 'Tool Fabric',
purpose: 'Standardized connectors to enterprise systems with input/output schemas, permission requirements, risk levels, approval requirements, and rollback options.',
principles: ['Act', 'Audit'] as PrincipleLabel[],
adjacent: ['Business Ontology', 'Human-in-the-Loop', 'Agent Runtime'],
category: 'Execution' as const,
},
{
id: 'human-in-the-loop',
name: 'Human-in-the-Loop',
purpose: 'Human approvals are first-class. Agents recommend; humans approve, edit, or reject. Approval gates pause execution and persist state for safe resumption.',
principles: ['Act', 'Verify'] as PrincipleLabel[],
adjacent: ['Tool Fabric', 'Durable Execution', 'Command Surface'],
category: 'Governance' as const,
},
{
id: 'durable-execution',
name: 'Durable Execution',
purpose: 'Workflows persist state, survive interruptions, and resume from the last checkpoint. Agents cannot lose progress across long-running tasks.',
principles: ['Act', 'Verify'] as PrincipleLabel[],
adjacent: ['Human-in-the-Loop', 'Agent Runtime', 'Evaluation'],
category: 'Execution' as const,
},
{
id: 'agent-roles',
name: 'Agent Roles',
purpose: 'Specialized agents — planner, research, execution, verifier, risk, narrative, audit — each with defined scope, tool access, and handoff contracts.',
principles: ['Plan', 'Act', 'Verify'] as PrincipleLabel[],
adjacent: ['Agent Runtime', 'Agent Identity', 'Orchestration'],
category: 'Core' as const,
},
{
id: 'evaluation',
name: 'Evaluation',
purpose: 'An evaluation harness that scores agents on accuracy, success rate, hallucination rate, approval compliance, cost, and outcome quality.',
principles: ['Verify', 'Learn'] as PrincipleLabel[],
adjacent: ['Durable Execution', 'Memory', 'Agent Runtime'],
category: 'Intelligence' as const,
},
{
id: 'memory',
name: 'Memory',
purpose: 'Layered memory: short-term task context, long-term preferences, business memory (accounts, vendors, policies), outcome memory, and institutional patterns.',
principles: ['Understand', 'Learn'] as PrincipleLabel[],
adjacent: ['Business Ontology', 'Evaluation', 'Agent Runtime'],
category: 'Intelligence' as const,
},
{
id: 'agent-identity',
name: 'Agent Identity',
purpose: 'Each agent has identity, role, permissions, allowed tools, risk level, approval threshold, and full audit trail — enforced at the platform layer.',
principles: ['Audit', 'Plan'] as PrincipleLabel[],
adjacent: ['Agent Roles', 'Agent Runtime', 'Tool Fabric'],
category: 'Governance' as const,
},
{
id: 'command-surface',
name: 'Command Surface',
purpose: 'The operator interface: signal timeline, agent queue, pending approvals, recommended actions, evidence panel, confidence scores, and execution logs.',
principles: ['Understand', 'Act', 'Audit'] as PrincipleLabel[],
adjacent: ['Human-in-the-Loop', 'Orchestration', 'Agent Runtime'],
category: 'Core' as const,
},
{
id: 'orchestration',
name: 'Orchestration',
purpose: 'Multi-agent coordination justified by cross-domain tasks or distinct security boundaries — specialization, scalability, maintainability, and optimization.',
principles: ['Plan', 'Act'] as PrincipleLabel[],
adjacent: ['Agent Runtime', 'Agent Roles', 'Command Surface'],
category: 'Execution' as const,
},
] as const;
export const IMPLEMENTATION_PRIORITIES = [
{ num: '01', name: 'Agent Registry', desc: 'Roles, permissions, and risk classification for every agent in the system.' },
{ num: '02', name: 'Tool Fabric', desc: 'Schemas, permissions, logging, and approval rules for every connector.' },
{ num: '03', name: 'Proof Chain', desc: 'Record every recommendation, source, tool call, approval, and result.' },
{ num: '04', name: 'Outcome Graph', desc: 'Link actions to real-world business outcomes for closed-loop learning.' },
{ num: '05', name: 'Human Approval Queue', desc: 'Approval, rejection, editing, escalation, and evidence requests.' },
{ num: '06', name: 'Verifier Agent', desc: 'Ensure tasks complete successfully before closing the proof chain.' },
{ num: '07', name: 'Evaluation Harness', desc: 'Regression testing and continuous monitoring for every agent.' },
{ num: '08', name: 'Command Surface UI', desc: 'Operator-facing interface for signal, approval, and execution management.' },
] as const;
export type ApplicationStatus = 'operational' | 'in-development' | 'planned';
export type ApplicationSector = 'Maritime' | 'Legal' | 'Real Estate' | 'Defense & Security' | 'Cyber' | 'Revenue' | 'Operations' | 'Consulting' | 'AI Infrastructure' | 'Infrastructure' | 'Crisis' | 'Intelligence';
export interface ApplicationEntry {
id: string;
name: string;
vertical: string;
sector: ApplicationSector;
status: ApplicationStatus;
tier: string;
description: string;
icon: string;
color: string;
principles: PrincipleLabel[];
registryProfileId?: string;
}
export const APPLICATIONS: ApplicationEntry[] = [
{
id: 'vessels', name: 'Vessels', vertical: 'Maritime', sector: 'Maritime',
status: 'operational', tier: 'Enterprise',
description: DOMAIN_PROFILES.vessels.description + ' AIS ingestion, voyage economics, port scheduling, demurrage risk, and sanctions screening.',
icon: '⚓', color: '#8a8a8a',
principles: ['Ingest', 'Understand', 'Audit'],
registryProfileId: 'vessels',
},
{
id: 'counsel', name: 'Counsel', vertical: 'Legal', sector: 'Legal',
status: 'operational', tier: 'Enterprise',
description: 'Legal matter lifecycle — filings, obligations, risk scoring, document intelligence, and privilege review automation with chain-of-custody proof.',
icon: '⚖', color: '#c9b787',
principles: ['Ingest', 'Plan', 'Audit'],
registryProfileId: 'prism',
},
{
id: 'terra', name: 'Terra', vertical: 'Real Estate', sector: 'Real Estate',
status: 'operational', tier: 'Enterprise',
description: DOMAIN_PROFILES.terra.description + ' Climate-adjusted valuations, deal pipeline analytics, tenant churn prediction, and ESG compliance reporting.',
icon: '▣', color: '#8a8a8a',
principles: ['Understand', 'Act', 'Verify'],
registryProfileId: 'terra',
},
{
id: 'aegis', name: 'Aegis', vertical: 'Defense & Security', sector: 'Defense & Security',
status: 'operational', tier: 'Sovereign',
description: DOMAIN_PROFILES.aegis.description + ' MITRE ATT&CK mapping, threat detection, incident response, and resilience verification.',
icon: '⬡', color: '#f5f5f5',
principles: ['Ingest', 'Verify', 'Audit'],
registryProfileId: 'aegis',
},
{
id: 'sentra', name: 'Sentra', vertical: 'Cyber Resilience', sector: 'Cyber',
status: 'operational', tier: 'Enterprise',
description: 'Cyber resilience command — posture monitoring, threat surface analysis, CISO intelligence, and incident escalation with governed approval flow.',
icon: '⬡', color: '#c9b787',
principles: ['Ingest', 'Understand', 'Verify'],
},
{
id: 'lyte', name: 'Lyte', vertical: 'Revenue Intelligence', sector: 'Revenue',
status: 'operational', tier: 'Enterprise',
description: DOMAIN_PROFILES.lyte.description + ' Pipeline health, forecast modeling, decision debt ledger, and governed deal actions with full attribution.',
icon: '◆', color: '#c9b787',
principles: ['Understand', 'Plan', 'Learn'],
registryProfileId: 'lyte',
},
{
id: 'pulse', name: 'Pulse', vertical: 'Founder Operations', sector: 'Operations',
status: 'operational', tier: 'Platform',
description: 'Founder operating channel — daily briefings, signal synthesis across all domains, and decision orchestration for executive operators.',
icon: '◉', color: '#8a8a8a',
principles: ['Understand', 'Act', 'Audit'],
},
{
id: 'carlota-jo', name: 'Carlota Jo', vertical: 'Consulting', sector: 'Consulting',
status: 'operational', tier: 'Professional',
description: DOMAIN_PROFILES.carlota.description + ' Advisory brief generation, engagement signal tracking, and proof-chained deliverables.',
icon: '◎', color: '#c9b787',
principles: ['Plan', 'Act', 'Learn'],
registryProfileId: 'carlota',
},
{
id: 'nuro-forge', name: 'NuroForge', vertical: 'AI Infrastructure', sector: 'AI Infrastructure',
status: 'in-development', tier: 'Platform',
description: 'Agent forge — custom agent training, fine-tuning orchestration, and evaluation harness for domain-specific model optimization.',
icon: '⬟', color: '#5e5e5e',
principles: ['Plan', 'Learn'],
},
{
id: 'meridian', name: 'Meridian', vertical: 'Infrastructure', sector: 'Infrastructure',
status: 'in-development', tier: 'Enterprise',
description: 'Infrastructure intelligence — cloud cost optimization, capacity planning, incident correlation, and governed change management.',
icon: '⬡', color: '#5e5e5e',
principles: ['Ingest', 'Plan', 'Verify'],
},
{
id: 'firestorm', name: 'Firestorm', vertical: 'Crisis Operations', sector: 'Crisis',
status: 'planned', tier: 'Sovereign',
description: 'Crisis operations — incident command, rapid response orchestration, impact simulation, and post-incident learning with full proof coverage.',
icon: '⬢', color: '#5e5e5e',
principles: ['Act', 'Verify', 'Learn'],
},
{
id: 'constellation', name: 'Constellation', vertical: 'Graph Intelligence', sector: 'Intelligence',
status: 'planned', tier: 'Platform',
description: 'Cross-domain intelligence graph — entity relationships, causal chains, emergent patterns, and cross-vertical signal correlation.',
icon: '✦', color: '#5e5e5e',
principles: ['Understand', 'Learn'],
},
];