huytieu/COG-second-brain

What it solves

COG is an agentic "second brain" designed to move beyond simple note-taking. It solves the problem of manual knowledge management by using AI agents to automatically classify braindumps, synthesize insights, and maintain a structured knowledge base without requiring a database or vendor lock-in.

How it works

COG operates as a set of instructions and skills provided to an AI agent (such as Claude Code, Cursor, or Gemini CLI) that interacts with a local directory of Markdown files. It uses a worker-agent architecture where specialized workers (e.g., data collectors, researchers) handle I/O tasks and a lead agent performs reasoning. To ensure accuracy, it employs a "Verification Harness" where read-only verifiers independently check artifacts against acceptance criteria.

Who it’s for

It is designed for individuals and teams—specifically Product Managers, Engineering Leads, and Founders—who need to automate their professional workflows, from strategic research and PRD generation to team intelligence briefs and people CRM tracking.

Highlights

• Self-Evolving Knowledge: Automatically learns patterns and consolidates scattered notes into frameworks.
• Agent-Agnostic: Works across multiple AI agents including Claude Code, Antigravity, Cursor, and Gemini CLI.
• Verification-First: Uses a V-model closed loop to ensure evidence-based analysis and prevent "AI slop."
• Local-First Privacy: Stores all data in plain Markdown files, supporting Git and iCloud sync for versioning and multi-device access.
• Specialized Role Packs: Provides tailored skill sets for different professional roles like PMs, Engineers, and Designers.