architecture-cortex-ledger

Define the architectural specification for the Theseus Database Cortex event-sourcing ledger.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/HugoAlmeidaMoreira/zeus-agent --skill architecture-cortex-ledger
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: architecture-cortex-ledger
Source: https://github.com/HugoAlmeidaMoreira/zeus-agent/tree/main/skills/architecture-cortex-ledger
Command: npx skills add https://github.com/HugoAlmeidaMoreira/zeus-agent --skill architecture-cortex-ledger

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The Theseus Database Cortex replaces static file-based memory with a continuous, event-driven ledger, enabling persistent, auditable memory streams across agents.

Core Features & Use Cases

  • Event-sourced memory ledger: stores a sequential timeline of all agent actions and external inputs.
  • Dossiers: group related events into thematic contexts for efficient retrieval.
  • Inter-agent inbox: lightweight pub/sub via PostgreSQL to coordinate tasks without heavy protocols.
  • Watchtower sync: automatic synchronization of local state to the central mothership database.
  • Ingestion pipeline: deterministic processing of events via dreaming agents that classify, summarize, and attach context.
  • Schema-driven: explicit definitions for events and dossiers to ensure consistency across agents.

Quick Start

Follow the Theseus Cortex setup guide at ~/office/knowledge/cortex/agent_cortex_setup.md to bootstrap and initialize the event-sourcing ledger in your environment.

Frequently Asked Questions about architecture-cortex-ledger

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How does an event-sourcing ledger for agent memory work?

An event-sourcing ledger replaces static file-based memory by storing a continuous, sequential timeline of all agent actions and external inputs as structured events. These events are grouped into dynamic dossiers for efficient retrieval and persistent, auditable memory streams.

Can I use PostgreSQL for inter-agent messaging and pub/sub?

Yes, PostgreSQL enables a lightweight inter-agent inbox workflow that coordinates tasks via pub/sub without heavy protocols. It acts as the central event-sourcing ledger, managing the continuous stream of structured events and dynamic dossiers for all agents.

What is the best way to bootstrap an event-sourced memory ledger for new agents?

Follow the setup guide at ~/office/knowledge/cortex/agent_cortex_setup.md to bootstrap new agents. This initializes the event-sourcing ledger in your environment and ensures schema parity for explicit event and dossier definitions across the system.

How do I synchronize local agent memory state to a central database?

The watchtower synchronization pipeline automatically syncs local agent state to the central mothership database. This ensures persistent memory streams remain consistent and auditable across all distributed agents utilizing the event-sourced ledger.

When do I need dynamic dossiers for event-sourced memory retrieval?

Dynamic dossiers are needed when grouping related events from the continuous timeline into thematic contexts. They enable efficient retrieval and organized processing of agent actions and external inputs stored within the event-sourcing ledger.