Azamat's AI Executive Assistant

A personal second brain and executive assistant workspace powered by Claude Code. This project provides persistent context, structured workflows, and reusable skills so that every conversation with Claude starts with full awareness of who I am, what I'm working on, and how I like to work.

How It Works

This isn't a traditional codebase -- it's a knowledge system. Claude Code reads these files automatically at the start of every session, giving it context about my background, priorities, projects, and preferences without me having to re-explain anything.

CLAUDE.md                        # Main brain file -- identity, imports, how everything connects
CLAUDE.local.md                  # Personal overrides (git-ignored)

Context Layer

Files that tell Claude who I am and what I'm focused on:

context/
  me.md                          # Profile -- role, location, strengths, motivations
  work.md                        # Business -- revenue streams, tech stack, tools
  team.md                        # Team structure and communication channels
  current-priorities.md          # Top priorities (update when focus shifts)
  goals.md                       # Quarterly goals and milestones (update each quarter)

Projects

Each active project gets a folder with a README summarizing scope, status, and key details:

projects/
  avicenna/                      # AI healthcare platform (React, Firebase, Gemini API)
  busa-speaking-club/            # Community platform for students in Korea
  livit-platform/                # Full-stack real estate marketplace (NestJS + Next.js)
  morning-cup/                   # Coffee e-commerce with real-time chat

Skills

Reusable workflows that Claude activates automatically when relevant:

.claude/skills/
  generating-proposals/          # Freelance client proposal workflow + template

See skills-backlog.md for workflows planned but not yet built.

Rules

Behavioral rules that shape how Claude communicates and writes code:

.claude/rules/
  communication-style.md         # Tone, formatting, writing preferences
  development-conventions.md     # Tech stack assumptions, coding approach

Decision Log

Append-only record of meaningful decisions:

decisions/log.md

References & Templates

references/sops/                 # Standard operating procedures
references/examples/             # Example outputs and style guides
templates/session-summary.md     # End-of-session closeout template

Archives

Nothing gets deleted -- outdated material moves to archives/.

Maintenance

Cadence	Action
Weekly	Nothing required. Auto-memory handles daily learnings.
Monthly	Glance at context/current-priorities.md. Update if focus has shifted.
Quarterly	Update context/goals.md with new goals and milestones.
As needed	Log decisions. Add references. Build new skills.

To make Claude remember something permanently: "Remember that I always prefer X."

Tech Stack (of the assistant itself)

This is a plain markdown project -- no dependencies, no build step. It's designed to be read by Claude Code and version-controlled with git.

• Claude Code -- AI assistant runtime
• Git -- Version control for all context and decisions
• Markdown -- Universal format for all files