Xbot77 Architecture Overview

Xbot77 is a production-grade AI Agent platform designed for real businesses. It combines AI Agents, an integrated CRM, governed workflows, and a ledger-authoritative token economy into a single, transparent system.

Core Architectural Principles

Server-authoritative logic — no client-side trust
Deterministic workflows — no hidden execution paths
Ledger-based accounting — every cost is traceable
Strict ownership boundaries — users, agents, and data are isolated
Fail-open ingestion — CRM and workflows never block users

Multi-Worker Cloudflare Architecture

Xbot77 is deployed as multiple Cloudflare Workers, each with a single responsibility:

Apex Worker — public website and documentation
User Worker — dashboards, agents, workflows, and UI
Admin Worker — platform governance and operations
CRM Worker — contacts, messages, identity, sessions
Storage Worker — R2-backed file and asset delivery

AI Agents

Agents are persistent, scoped entities with knowledge, rules, memory, and cost controls. Each agent operates within explicit boundaries and consumes tokens only through server-authoritative execution paths.

Workflow Engine

Workflows are deterministic graphs composed of typed nodes (forms, emails, logic). Each execution is assigned a single authoritative execution session ID that governs:

Token accounting
Ledger entries
CRM side effects
UI reporting

Token Economy & Ledger

All AI usage and automation costs are recorded in an append-only token ledger. Balances are never inferred or reconstructed. Every charge is auditable, execution-scoped, and immutable.

Why This Matters

Xbot77 avoids black-box AI behavior. Businesses can trust costs, behavior, and data ownership — making the platform suitable for real operations, not just demos.