Spark - Secure Personal AI Research Kit

Spark is a powerful, multi-provider LLM interface for conversational AI with integrated tool support. It supports AWS Bedrock, Anthropic Direct API, and Ollama local models through both CLI and Web interfaces.

Key Features

Multi-Provider Support - AWS Bedrock, Anthropic Direct API, and Ollama local models
Dual Interface - Rich CLI terminal UI and modern Web browser interface
MCP Tool Integration - Connect external tools via Model Context Protocol
Intelligent Context Management - Automatic conversation compaction with model-aware limits
Security Features - Prompt inspection, tool permissions, and audit logging
Multiple Database Backends - SQLite, MySQL, PostgreSQL, and Microsoft SQL Server

Quick Start

Installation

pip install dtSpark

First-Time Setup

Run the interactive setup wizard to configure Spark:

spark --setup

This guides you through:

LLM provider selection and configuration
Database setup
Interface preferences
Security settings

Running Spark

# Start with CLI interface
spark

# Or use the alternative command
dtSpark

Documentation

Comprehensive documentation is available in the docs folder:

Installation Guide - Detailed installation instructions
Configuration Reference - Complete config.yaml documentation
Features Guide - Detailed feature documentation
CLI Reference - Command-line options and chat commands
Web Interface - Web UI guide
MCP Integration - Tool integration documentation
Security - Security features and best practices

Architecture Overview

graph LR
    subgraph Interfaces
        CLI[CLI]
        WEB[Web]
    end

    subgraph Core
        CM[Conversation<br/>Manager]
    end

    subgraph Providers
        BEDROCK[AWS Bedrock]
        ANTHROPIC[Anthropic]
        OLLAMA[Ollama]
    end

    subgraph Tools
        MCP[MCP Servers]
        BUILTIN[Built-in Tools]
    end

    CLI --> CM
    WEB --> CM
    CM --> BEDROCK
    CM --> ANTHROPIC
    CM --> OLLAMA
    CM --> MCP
    CM --> BUILTIN

Requirements

Python 3.10 or higher
AWS credentials (for Bedrock)
Anthropic API key (for direct API)
Ollama server (for local models)

Licence

MIT Licence - see LICENSE for details.

Author

Matthew Westwood-Hill matthew@digital-thought.org

Support

Documentation: docs/
Issues: GitHub Issues

Release files for dtSpark 1.0.11

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for dtSpark 1.0.11
File	Size	Uploaded
dtspark-1.0.11.tar.gz	315.2 kB	Jan 27, 2026	Details

Built distribution (wheel)

Table of built distributions (wheels) for dtSpark 1.0.11
File	Interpreter	ABI	Platform	Reset
dtspark-1.0.11-py3-none-any.whl 331.4 kB Jan 27, 2026	Python 3	none	any	Details

Total release size: 646.5 kB

Release files / dtspark-1.0.11.tar.gz

Download URL	dtspark-1.0.11.tar.gz
Size	315.2 kB
Tags	Source
SHA-256 checksum How to use checksums	`0f6ce8544cb4e9c6194411589e671a1d7b4e78ec930c406858f3a66e66c82cfc`
BLAKE2b-256 checksum How to use checksums	`bb4d32791b2f01053d33a448e17eb17201a39b732472667ed44d1b2d59a0a21b`
Upload date	Jan 27, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	Yes
Uploaded via	`twine/6.1.0 CPython/3.13.7`

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Identity

Publishing platform: GitHub Actions
Publishing repository: github.com/Digital-Thought/dtSpark
Publishing commit: github.com/Digital-Thought/dtSpark/tree/f3c1cd8cb686f4a0e44619c578f754f9ea489e4a

Workflow

Publishing configuration: github.com/Digital-Thought/dtSpark/blob/f3c1cd8cb686f4a0e44619c578f754f9ea489e4a/.github/workflows/publish-to-pypi.yml
Publishing logs: github.com/Digital-Thought/dtSpark/actions/runs/21392183261/attempts/1

Artifact

Subject: dtspark-1.0.11.tar.gz
SHA-256 checksum: 0f6ce8544cb4e9c6194411589e671a1d7b4e78ec930c406858f3a66e66c82cfc Verifying attestations

Release files / dtspark-1.0.11-py3-none-any.whl

Download URL	dtspark-1.0.11-py3-none-any.whl
Size	331.4 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`75414471745e5a946b40b4cd0ea682a614c6805cc4f051c8586426863c127d6f`
BLAKE2b-256 checksum How to use checksums	`8ed5fcc266a35feb522163104a0684b0b7adc204bf7e981da6b6830d16021c14`
Upload date	Jan 27, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	Yes
Uploaded via	`twine/6.1.0 CPython/3.13.7`

Provenance

Identity

Publishing platform: GitHub Actions
Publishing repository: github.com/Digital-Thought/dtSpark
Publishing commit: github.com/Digital-Thought/dtSpark/tree/f3c1cd8cb686f4a0e44619c578f754f9ea489e4a

Workflow

Publishing configuration: github.com/Digital-Thought/dtSpark/blob/f3c1cd8cb686f4a0e44619c578f754f9ea489e4a/.github/workflows/publish-to-pypi.yml
Publishing logs: github.com/Digital-Thought/dtSpark/actions/runs/21392183261/attempts/1

Artifact

Subject: dtspark-1.0.11-py3-none-any.whl
SHA-256 checksum: 75414471745e5a946b40b4cd0ea682a614c6805cc4f051c8586426863c127d6f Verifying attestations