dredge-cli 0.1.4

DREDGE — A small Python utility (scaffold).

DREDGE

DREDGE — small Python package scaffold.

Repository Structure

• src/dredge/ - Python package source code
• tests/ - Test files
• docs/ - Documentation files
• benchmarks/ - Benchmark scripts and results
• swift/ - Swift implementation files
• DREDGE-Cli.xcworkspace - Xcode workspace for Swift development
• archives/ - Archived files (excluded from version control)

Install

Create a virtual environment and install:

python -m venv .venv source .venv/bin/activate # or .venv\Scripts\activate on Windows pip install -e .

Server Usage

DREDGE includes two web servers:

1. DREDGE x Dolly Server (Port 3001)

A web server for API-based interaction with Dolly integration.

Starting the Server

python -m dredge serve
# or
dredge-cli serve --host 0.0.0.0 --port 3001 --debug

API Endpoints

• GET / - API information and available endpoints
• GET /health - Health check endpoint
• POST /lift - Lift an insight with Dolly integration

Example Usage

# Get API info
curl http://localhost:3001/

# Check health
curl http://localhost:3001/health

# Lift an insight
curl -X POST http://localhost:3001/lift \
  -H "Content-Type: application/json" \
  -d '{"insight_text": "Digital memory must be human-reachable."}'

2. MCP Server (Port 3002) - Quasimoto Integration

A Model Context Protocol server for serving Quasimoto neural wave function models.

Starting the MCP Server

python -m dredge mcp
# or
dredge-cli mcp --host 0.0.0.0 --port 3002 --debug

MCP Protocol Endpoint

• GET / - MCP server capabilities and model information
• POST /mcp - MCP protocol endpoint for model operations

Available Operations

1. list_capabilities - List available models and operations
2. load_model - Load Quasimoto models (1D, 4D, 6D, ensemble)
3. inference - Run inference on loaded models
4. get_parameters - Retrieve model parameters
5. benchmark - Run performance benchmarks

Example MCP Request

# List capabilities
curl http://localhost:3002/

# Load a model
curl -X POST http://localhost:3002/mcp \
  -H "Content-Type: application/json" \
  -d '{"operation": "load_model", "params": {"model_type": "quasimoto_1d"}}'

# Run inference
curl -X POST http://localhost:3002/mcp \
  -H "Content-Type: application/json" \
  -d '{"operation": "inference", "params": {"model_id": "quasimoto_1d_0", "inputs": {"x": [0.5], "t": [0.0]}}}'

Available Models

• quasimoto_1d - 1D wave function (8 parameters)
• quasimoto_4d - 4D spatiotemporal wave function (13 parameters)
• quasimoto_6d - 6D high-dimensional wave function (17 parameters)
• quasimoto_ensemble - Configurable ensemble models

GitHub Codespaces

The repository includes .devcontainer/devcontainer.json configured to automatically forward ports 3001 and 3002 when running in GitHub Codespaces.

Swift Development

DREDGE includes a Swift CLI implementation. You can develop using:

Xcode Workspace

open DREDGE-Cli.xcworkspace

Swift Package Manager

# Build from root
swift build
swift run dredge-cli

# Or build from swift/ directory
cd swift
swift build
swift run dredge-cli

See SWIFT_PACKAGE_GUIDE.md for detailed Swift development information.

Test

Run tests with pytest:

pip install -U pytest pytest

Development

• Edit code in src/dredge
• Update version in pyproject.toml
• Tag releases with v and push tags