INTENT-ITERATIVE

DSL-based intent execution system with iterative refinement, AMEN boundary, and AI-powered assistance

Overview

INTENT-ITERATIVE is a system that allows you to:

1. Define intents using a simple YAML-based DSL
2. Simulate execution with dry-run planning
3. Get AI suggestions using local LLMs via Ollama
4. Iteratively refine your intent through feedback loops
5. Execute safely with the AMEN boundary (explicit approval required)

Architecture

┌─────────────────────┐
│  CLI / Web UI       │  ← User interface
└─────────┬───────────┘
          ↓
┌─────────────────────┐
│ Parser / Validator  │  ← DSL → IR
└─────────┬───────────┘
          ↓
┌─────────────────────┐
│ Intermediate Rep.   │  ← Canonical state
└─────────┬───────────┘
          ↓
┌─────────────────────┐
│ Planner / Simulator │  ← Dry-run
└─────────┬───────────┘
          ↓
┌─────────────────────┐
│ AI Gateway (Ollama) │  ← LLM suggestions
└─────────┬───────────┘
          ↓
┌─────────────────────┐
│ Feedback Loop       │  ← Iterate
└─────────┬───────────┘
          ↓
┌─────────────────────┐
│ AMEN Boundary       │  ← Explicit approval
└─────────┬───────────┘
          ↓
┌─────────────────────┐
│ Executor            │  ← Real execution
└─────────────────────┘

Quick Start

Installation

# Clone repository
git clone https://github.com/softreck/intent-iterative.git
cd intent-iterative

# Full setup (recommended)
make setup

# Or manual install
pip install -r requirements.txt
pip install litellm
cp .env.example .env

Configuration (.env)

Copy .env.example to .env and adjust:

# Server
HOST=0.0.0.0
PORT=8080

# AI Gateway
OLLAMA_BASE_URL=http://localhost:11434
DEFAULT_MODEL=llama3.2
MAX_MODEL_PARAMS=12.0

# Execution (no AMEN prompt by default)
SKIP_AMEN_CONFIRMATION=true
CONTAINER_PORT=8000

AI Gateway Setup (Ollama)

# Install Ollama
curl -fsSL https://ollama.com/install.sh | sh

# Start and pull model
make ollama-start
make ollama-pull

# Or manually
ollama serve
ollama pull llama3.2

Using Makefile

make help          # Show all commands
make setup         # Full setup
make web           # Start web server
make shell         # Interactive shell
make execute       # Execute example intent
make test          # Run all tests
make ollama-models # List available models
make clean         # Clean temp files

Shell Interface

# Start interactive shell
make shell
# Or: python -m cli.main

# Execute intent directly (no AMEN prompt)
make execute
# Or: python -m cli.main execute examples/user-api.intent.yaml

Interactive Shell Commands:

intent> new my-api          # Create new intent
intent> load file.yaml      # Load from file
intent> plan                # Run dry-run
intent> suggest             # Get AI suggestions
intent> apply               # Auto-apply AI suggestions
intent> chat                # Chat with AI
intent> iterate             # Apply manual changes
intent> amen                # Approve execution
intent> execute             # Execute approved intent
intent> show [json]         # Show current state
intent> models              # List AI models
intent> ai-health           # Check AI Gateway status
intent> help                # Show help
intent> exit                # Exit shell

Web Interface

# Start web server
python -m web.app

# Open browser at http://localhost:8080

AI Gateway

The AI Gateway uses LiteLLM to provide unified access to local LLMs via Ollama.

Supported Models (≤12B parameters)

Model	Size	Description
llama3.2	3B	Default - Fast and efficient
llama3.2:1b	1B	Ultra lightweight
llama3.1:8b	8B	Balanced performance
mistral	7B	Fast inference
mistral-nemo	12B	Best quality under 12B
gemma2	9B	Google Gemma 2
gemma2:2b	2B	Lightweight
phi3	3.8B	Microsoft Phi-3
qwen2.5	7B	Alibaba Qwen 2.5
codellama	7B	Code generation
codegemma	7B	Google CodeGemma
deepseek-coder	6.7B	DeepSeek Coder

Configuration

Environment variables:

export OLLAMA_BASE_URL="http://localhost:11434"
export DEFAULT_MODEL="llama3.2"
export MAX_MODEL_PARAMS="12.0"

DSL Format

INTENT:
  name: user-api
  goal: Create a REST API for user management

ENVIRONMENT:
  runtime: docker
  base_image: python:3.12-slim
  ports:
    - 8000

IMPLEMENTATION:
  language: python
  framework: fastapi
  actions:
    - api.expose GET /ping
    - api.expose GET /users
    - api.expose POST /users
    - api.expose DELETE /users/{id}

EXECUTION:
  mode: dry-run

Supported Actions

Action	Format	Description
api.expose	api.expose METHOD /path	Expose HTTP endpoint
db.create	db.create table_name	Create database table
db.add_column	db.add_column table column type	Add column to table
shell.exec	shell.exec command	Execute shell command
rest.call	rest.call METHOD url	Call external REST API
file.create	file.create path	Create file

API Reference

REST Endpoints

Method	Endpoint	Description
GET	/api/intents	List all intents
POST	/api/intents/parse	Parse DSL and create intent
GET	/api/intents/{id}	Get intent by ID
DELETE	/api/intents/{id}	Delete intent
POST	/api/intents/{id}/plan	Run dry-run
POST	/api/intents/{id}/iterate	Apply changes
POST	/api/intents/{id}/amen	Approve for execution
POST	/api/intents/{id}/execute	Execute approved intent
GET	/api/intents/{id}/code	Get generated code

AI Gateway Endpoints

Method	Endpoint	Description
GET	/api/ai/status	Check AI Gateway status
GET	/api/ai/models	List available models
POST	/api/ai/complete	Generate AI completion
POST	/api/ai/chat	Chat with AI
POST	/api/intents/{id}/ai/suggest	Get AI suggestions
POST	/api/intents/{id}/ai/apply	Auto-apply suggestions

Python API

from ir.models import IntentIR
from parser import parse_dsl
from planner import plan_intent
from ai_gateway import get_gateway, create_feedback_loop

# Parse DSL
ir = parse_dsl(dsl_content)

# Run dry-run
result = plan_intent(ir)
print(result.generated_code)

# Get AI suggestions
loop = create_feedback_loop()
suggestions = loop.analyze(ir, focus="security")
print(suggestions.suggestions)

# Apply AI suggestions
loop.iterate(ir, auto_apply=True)

# Or chat directly
gateway = get_gateway()
response = gateway.complete("Explain FastAPI middleware")
print(response["content"])

Testing

# Run all tests (58 tests)
pytest

# Run specific test suites
pytest tests/e2e/test_shell.py -v      # 17 tests
pytest tests/e2e/test_web.py -v        # 18 tests  
pytest tests/e2e/test_ai_gateway.py -v # 23 tests

# Run with coverage
pytest --cov=. --cov-report=html

Project Structure

intent-iterative/
├── ir/                 # Intermediate Representation models
├── parser/             # DSL parser
├── planner/            # Dry-run simulator
├── executor/           # Execution engine
├── ai_gateway/         # LiteLLM AI Gateway
│   ├── gateway.py      # Main gateway with Ollama models
│   └── feedback_loop.py # LLM-powered feedback loop
├── cli/                # Shell interface
├── web/                # Web interface (FastAPI)
├── tests/e2e/          # E2E test suite
│   ├── test_shell.py
│   ├── test_web.py
│   └── test_ai_gateway.py
├── examples/           # Example DSL files
├── config.py           # Configuration loader
├── Makefile            # Build automation
├── .env.example        # Environment template
├── requirements.txt
└── README.md

Workflow

1. Define Intent → Write DSL or use web editor
2. Parse → Convert DSL to IR (Intermediate Representation)
3. Plan → Run dry-run simulation, review generated code
4. Get AI Suggestions → Analyze with local LLM
5. Iterate → Make changes, re-plan until satisfied
6. AMEN → Explicitly approve for execution
7. Execute → Run with real side effects (Docker build, etc.)

License

Apache 2 License - see LICENSE file.