Skip to content

The local-first relational and vector database engine for Cortex-Works. Built with SQLite and LanceDB, it brings Agile methodologies and Scrum ticket management directly to autonomous multi-agent teams.

License

Notifications You must be signed in to change notification settings

cortex-works/cortex-db

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

6 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

CortexDB

The New Brain: Local-first Database Engine for CortexSync Ecosystem

Warning

🚧 Alpha Version (Under Active Development) This project is currently in early alpha stage and is undergoing heavy development. Features, APIs, and CLI commands are subject to breaking changes without notice.

Current Testing Environment (As of 25 Feb 2026): Testing is exclusively performed on a developer machine running macOS Sequoia 15.7.4 with the latest versions of compatible IDEs (Cursor/VS Code/Windsurf).

If you are cloning this repository to run on Windows, Linux, or older IDE versions, you may encounter instability or broken features. Cross-platform testing will be expanded in future releases.

🌟 Overview

CortexDB is the dedicated database engine for the Cortex-Works ecosystem, managing Scrum logic (Epics, Tickets, Backlog) and Agent Memory. Designed specifically to transition away from flat JSONL ledgers to relational and vector storage, allowing for scalable, enterprise-grade AI agent tracking.

πŸ›οΈ The "Data Triad" Architecture

CortexDB plays a central role in the Data Triad Architecture:

  1. CortexSync: The Pulsar / Event Router (Watchers and notification endpoints).
  2. CortexDB (This Repo): The New Brain. Manages state and database integration.
    • SQLite (Local): Stores relational data like Epic/Ticket tracking, assignments, and structural relationships.
    • LanceDB (Local): Lightweight vector database for semantic search and Retrieval-Augmented Generation (RAG).
  3. CortexRelay: The Sync & Backup Engine.

πŸ’‘ Why CortexDB?

  • Performance: Blazing fast SQLite for querying and LanceDB for vector search in milliseconds.
  • Scale: Seamlessly supports UI integration for Kanban boards, burn-down charts, and sprint backlogs in cortex-studio.
  • Separation of Concerns: DB logic is decoupled from watcher threads in cortex-sync, avoiding "God Object" congestion and conflict.

πŸš€ Phase 3 & 4 Roadmap

This repository powers Phase 3 of the Cortex-Works Ecosystem roadmap:

  • Migration from JSONL to Local-First DBs: Decouple the Data Layer from the Sync Engine.
  • SQLite Integration: Store relational data to prepare for the workflow structure system (Epic, Ticket, Agent Assignment).
  • Embedded LanceDB: Isolate Vector Storage for millisecond semantic search (Zero-config).

About

The local-first relational and vector database engine for Cortex-Works. Built with SQLite and LanceDB, it brings Agile methodologies and Scrum ticket management directly to autonomous multi-agent teams.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages