Skip to content

Releases: KrystianYCSilva/itzamna-prompt-os

Adicionando documentação e Ajustes finos

11 Feb 17:46
@KrystianYCSilva KrystianYCSilva
1.0.1
a17ed1c

Choose a tag to compare

View all tags
Adicionando documentação e Ajustes finos Latest
Latest

Adicionando documentação e Ajustes finos

Assets 2
Loading

Versão oficial 1.0.0

11 Feb 17:33
@KrystianYCSilva KrystianYCSilva
1.0.0
7f4e4f0

Choose a tag to compare

View all tags
Versão oficial 1.0.0

Criando a primeira versão do Projeto!

Itzamna PromptOS

Cognitive orchestrator for AI coding agents

Version
CoALA
License

What is Itzamna?

Itzamna teaches AI coding agents to think with context, memory, and workflows. It's a cognitive orchestrator that transforms basic AI CLIs into intelligent, memory-aware development assistants.

Key capabilities:

  • Python CLI (itzamna): Bootstrap, status, check commands
  • Cognitive Kernel: 3-level decision system (K1 reflexive, K2 deliberate, K3 deep)
  • Memory System: CoALA-inspired 4-layer memory architecture
  • Context Engineering: Lean (3 files) or enterprise (10+ files) structures
  • AI Slash Commands (/itzamna.*): Memory & context management for agents

Named after the Mayan god of wisdom and writing, Itzamna brings intelligence to AI development workflows.

Architecture

PYTHON CLI (itzamna)  →  Bootstrap + Detection
                         ↓
                    Templates (kernel, memory, context)
                         ↓
    AI AGENT + KERNEL (K1/K2/K3)  →  MEMORY + CONTEXT
                         ↑
                    AGENTS.md
                    CONSTITUTION.md
                    MEMORY.md
  • Python CLI: Bootstrap, multi-CLI detection, status (typer + rich)
  • Cognitive Kernel: 3-level decision making (reflexive → deliberate → deep)
  • Memory: CoALA 4-layer architecture (Working, Project Context, Episodic, Semantic)
  • Multi-CLI: Auto-detects and installs to 17 AI CLIs simultaneously
  • Token Efficient: 95% reduction from v2 (~562 tokens vs ~12K+)

Supported CLIs (17 total)

CLI Commands Directory Core Files Type
GitHub Copilot .github/agents/ kernel.md, AGENTS.md IDE
Claude Code .claude/commands/ kernel.md, AGENTS.md CLI
Gemini CLI .gemini/commands/ GEMINI.md, kernel.md CLI
Cursor .cursor/commands/ kernel.md, AGENTS.md IDE
Qwen Code .qwen/commands/ kernel.md, AGENTS.md CLI
opencode .opencode/commands/ kernel.md, AGENTS.md CLI
Codex CLI .codex/prompts/ kernel.md, AGENTS.md CLI
Windsurf .windsurf/commands/ kernel.md, AGENTS.md IDE
Kilo Code .kilocode/commands/ kernel.md, AGENTS.md IDE
Auggie CLI .augment/commands/ kernel.md, AGENTS.md CLI
CodeBuddy .codebuddy/commands/ kernel.md, AGENTS.md CLI
Qoder CLI .qoder/commands/ kernel.md, AGENTS.md CLI
Roo Code .roo/commands/ kernel.md, AGENTS.md IDE
Amazon Q CLI .amazonq/commands/ kernel.md, AGENTS.md CLI
Amp .agents/ kernel.md, AGENTS.md CLI
SHAI .shai/commands/ kernel.md, AGENTS.md CLI
IBM Bob .bob/commands/ kernel.md, AGENTS.md IDE

Commands

Python CLI

Command Description
itzamna init Bootstrap: detect CLIs, install kernel, create MEMORY.md, setup .context/
itzamna check Show status (version, detected CLIs, installed commands)
itzamna version Show Itzamna CLI version
itzamna init --ai <cli> Bootstrap for specific CLI only
itzamna init /path/to/project Bootstrap in specific directory

AI Slash Commands

Command Description Purpose
/itzamna.init Verify Itzamna installation Check if kernel, memory, context are loaded
/itzamna.status System status Show CLI info, memory state, context structure
/itzamna.memory Memory management Read/update/check MEMORY.md (episodic memory)
/itzamna.context Context management Manage .context/ directory (status/check/upgrade/update)

Quick Start

1. Install Python CLI

uv tool install itzamna-cli --from git+https://github.com/KrystianYCSilva/itzamna-prompt-os.git

2. Initialize in Your Project

cd your-project/
itzamna init        # Detects CLIs, creates kernel, MEMORY.md, .context/
itzamna check       # Verify installation

3. Use Slash Commands (in AI CLI)

# Verify installation
/itzamna.init

# Check system status
/itzamna.status

# Read memory
/itzamna.memory read

# Check context structure
/itzamna.context status

4. Let the Agent Think with Kernel

The agent now has:

  • K1 (Reflexive): Fast decisions from MEMORY.md
  • K2 (Deliberate): Structured reasoning with .context/
  • K3 (Deep): Long-term planning with CONSTITUTION.md

How Itzamna Works

Phase 1: Detection    → Scan for AI CLI directories
Phase 2: Bootstrap    → Install kernel.md + AGENTS.md per CLI
Phase 3: Memory       → Create MEMORY.md (episodic memory template)
Phase 4: Context      → Setup .context/ (lean or enterprise)
Phase 5: Core         → Install WORKFLOWS.md, QUALITY-GATES.md
Phase 6: Commands     → Install 4 slash commands per CLI

Installation

Option 1: Python CLI (Recommended)

# Install globally with uv
uv tool install itzamna-cli --from git+https://github.com/KrystianYCSilva/itzamna-prompt-os.git

# Or with pipx
pipx install git+https://github.com/KrystianYCSilva/itzamna-prompt-os.git

# Or from local clone
git clone https://github.com/KrystianYCSilva/itzamna-prompt-os.git
cd itzamna-prompt-os
uv tool install .

See INSTALL.md for detailed installation instructions and troubleshooting.

Option 2: From Source

# Clone and install in development mode
git clone https://github.com/KrystianYCSilva/itzamna-prompt-os.git
cd itzamna-prompt-os
pip install -e .

# Or with uv
uv tool install --editable .

Verify Installation

itzamna version     # Check CLI version
itzamna check       # Check detected CLIs
/itzamna.init       # Check installation in AI CLI

What itzamna init Injects

For each detected AI CLI:

Core System Files

  1. kernel.md - 3-level cognitive decision system (K1/K2/K3)
  2. AGENTS.md - Entry point with rules and structure
  3. CONSTITUTION.md - T0/T1/T2 rule hierarchy
  4. MEMORY.md - Episodic memory template (CoALA-inspired)

Slash Commands (4 per CLI)

  • /itzamna.init - Verify installation
  • /itzamna.status - System status
  • /itzamna.memory - Read/update/check MEMORY.md
  • /itzamna.context - Manage .context/ directory

Workflows & Quality

  1. WORKFLOWS.md - Task workflows
  2. QUALITY-GATES.md - Quality criteria

Context Structure (Optional)

  1. .context/ - Lean (3 files) or enterprise (10+ files)
    • project.md - Project overview
    • tech.md - Tech stack
    • rules.md - Project rules
    • Enterprise: _meta/, standards/, patterns/, knowledge/, workflows/
Loading