AI Context System

Session Continuity

Resume exactly where you left off—days, weeks, or months later—without re-explaining anything.

Externalized Context

Make AI thoughts, decisions, and mental models visible to humans in structured documentation.

Human-AI Collaboration

Full visibility into AI reasoning enables programmers to work alongside AI agents effectively.

AI-to-AI Collaboration

Facilitate peer review, project handoffs, and collaborative development between AI agents.

Fast & Simple

Quick saves (2-3 min) for daily work. Comprehensive saves (10-15 min) before breaks.

Multi-AI Support

Universal file structure works with Claude Code, Cursor, Aider, GitHub Codex, and more.

Quick Start

bash

# One-command install
curl -sL https://raw.githubusercontent.com/rexkirshner/ai-context-system/main/install.sh | bash

# Initialize context
/init-context

# Daily workflow
/review-context  # Start of session (30 sec)
/save            # Quick update (2-3 min)
/save-full       # Comprehensive save (10-15 min)

What Gets Created

your-project/
├── CLAUDE.md                    # ← Auto-loaded by Claude Code
├── context/                     # Externalized AI context
│   ├── CONTEXT.md               # Project orientation
│   ├── STATUS.md                # Current state + Quick Reference
│   ├── DECISIONS.md             # Decision log (WHY)
│   └── SESSIONS.md              # History + mental models
├── .claude/
│   ├── commands/                # 22 slash commands
│   ├── agents/                  # 12 specialist agents (v5.0)
│   ├── schemas/                 # 7 JSON validation schemas
│   └── hooks/                   # Session automation
└── scripts/                     # Helper utilities

Core Philosophy

AI reasoning is invisible. Context is lost between sessions. Decisions lack rationale. Humans can't review AI's thinking.

The solution: Externalize AI context into structured, visible documentation that serves four purposes:

Session Continuity - Pick up exactly where you left off

Externalized Context - Make AI reasoning visible to humans

Human-AI Collaboration - Full visibility into AI thinking

AI-to-AI Collaboration - Perfect handoffs and peer review

Current Version

v5.0.2 - Agent-Based Code Review Architecture

Major release introducing self-declaring specialist agents with formal contracts:

12 Specialist Agents - Code Reviewer (orchestrator) + 8 Specialist Reviewers (Security, Performance, Accessibility, SEO, Database, Infrastructure, TypeScript, Testing) + 3 Support Agents (Codebase Scanner, Synthesis Agent, Audit Compare)

22 Slash Commands - Full command suite including code review orchestrator, specialized audits, context management, and session workflows

Self-Declaring Contracts - Agents declare their own capabilities via JSON Schema-validated contracts

Session-Start Hooks - Automatic context health checks at session start

Scope Boundaries - Clear ownership between agents preventing duplicate findings

Shell Compatibility - Full support for bash, zsh, and sh across macOS and Linux

80 tests passing - Comprehensive coverage including install, hooks, shell compatibility, and cross-platform

v4.x introduced modular code review, documentation health checking, and Quick Reference auto-generation.

AI Context SystemExternalize AI reasoning. Enable perfect session continuity.