laida-cli 0.1.0b5

LAIDA v3 - Agent-executable software architecture methodology CLI

LAIDA CLI

LLM Agent-Executable Implementation Design Architecture

A methodology-driven CLI that transforms project requirements into production-ready software through structured AI collaboration.

Key Features

• Claude Code Native - Built specifically for Claude Code, leveraging its agentic capabilities for autonomous software development
• Structured Methodology - 5 phases, 20 steps that guide from requirements to deployment
• Mission Control TUI - Interactive dashboard with real-time streaming output
• YAML-First Architecture - All artifacts are structured, traceable, and version-controllable
• Scale-Adaptive - Automatically adjusts methodology depth based on project complexity

Installation

# With TUI (Recommended)
pip install "laida-cli[tui]"

# Basic installation
pip install laida-cli

Requirements

• Python 3.11+
• Claude Code CLI installed and authenticated

Quick Start

# Initialize project structure
laida init

# Launch Mission Control
laida tui

Then use keyboard shortcuts:

• Press E to edit your project context (13-section questionnaire)
• Press L to authenticate with Claude Code
• Press R to run the next step

Mission Control TUI

╭─────────────────────────────────────╮
│  LAIDA Mission Control  β0.1.5     │
╰─────────────────────────────────────╯

An interactive terminal dashboard for orchestrating the LAIDA methodology:

Key	Action	Description
R	Run	Execute next phase/step with confirmation dialog
E	Edit	Open project_context.yaml in your editor
S	Status	Display project progress overview
L	Login	Authenticate with Claude Code
V	Validate	Check artifact integrity
Q	Quit	Exit Mission Control

Dashboard Components

• Live Feed - Real-time streaming output from Claude Code execution
• Progress Panel - Visual tracker showing completed/pending phases
• File Tree - Browse generated artifacts in outputs/
• Git Status - Current branch and working tree status
• Telemetry - Session time, task duration, memory usage

Commands

Command	Description
laida init	Initialize project with methodology templates
laida tui	Launch Mission Control dashboard
laida login	Authenticate with Claude Code
laida status	Show project progress
laida run <phase> <step>	Execute a specific step
laida validate	Validate generated artifacts

Methodology

LAIDA executes a structured 5-phase methodology:

Phase	Name	Deliverables
0	Strategic Foundation	Project scale, tech stack (TDR), risks, QAS, bounded contexts, ADRs
1	Module Design	RTM, module specifications, DDD aggregates
2	Physical Design	File structure, Chain-of-Thought architecture
3	Contract Definition	API contracts, dependency analysis
4	Implementation	Production code, tests, CI/CD, documentation

Each phase contains 4 steps, totaling 20 sequential steps from requirements to deployment.

Project Structure

your-project/
├── .laida/                              # Methodology engine
│   ├── config.yaml                      # Phase/step definitions
│   ├── prompts/                         # Claude Code prompts per step
│   ├── templates/                       # Artifact schemas
│   │   └── project_context_template.yaml
│   ├── matrices/                        # Tech selection criteria
│   └── rules/                           # Scale calculation
├── inputs/
│   ├── project_context.yaml             # YOUR PROJECT DETAILS
│   └── project_context_template.yaml    # Reference template
└── outputs/
    └── phase{0-4}/                      # Generated artifacts

Project Context

The project_context.yaml is your project's DNA. It contains 13 sections that LAIDA uses to make intelligent decisions:

Section	Purpose
Project Identity	Name, type, description
Business Context	Domain, criticality, deadlines
Scale Expectations	Users, RPS, data volume, growth
Team Context	Size, skills, availability
Technical Context	Infrastructure, legacy systems, integrations
Security	Auth, encryption, compliance (GDPR, HIPAA, etc.)
Observability	Logging, metrics, tracing, alerting
Recovery	RTO, RPO, backup strategy
Quality Standards	Coverage, complexity limits, performance budgets
Architecture Preferences	Paradigm, deployment model
Domain-Specific	Custom fields for your industry
Known Concerns	Documented risks and constraints
Success Criteria	Measurable goals

How It Works

1. You describe your project in project_context.yaml
2. LAIDA calculates project scale (small/medium/large) and selects optimal tech stack
3. Claude Code executes each step autonomously, generating structured YAML artifacts
4. Artifacts chain together - each step references outputs from previous steps
5. Phase 4 generates production Python code, tests, and CI/CD configuration

Development

git clone https://github.com/your-org/laida-cli
cd laida-cli
pip install -e ".[all]"

# Quality checks
pytest tests/ -v
ruff check src/
mypy src/laida_cli/

License

MIT

---

Built for Claude Code | Documentation | Issues