Local-first software & systems engineering
Practical tools.
Local control.
Keep your AI conversations, credentials, and working files useful—and under your control.
Distributed Thinking Systems builds local-first software and automation for developers, independent professionals, and small teams. Explore the tools, follow the engineering, or get help making a sensitive workflow work on your own systems.
Tools worth a closer look
Different jobs, a shared approach: local processing, portable files, and deliberate handling of sensitive information.
Safari extension
LogGPT and LogGPT Plus
Keep a local archive of your ChatGPT conversations. LogGPT Plus adds available uploaded and generated files in a portable ZIP, ready for backup or further processing.
Explore LogGPT →
Local audit and cleanup
UnicodeFix
Find hidden Unicode and inspect provenance and configured watermark signals in text and images. Review the evidence, then choose what to clean up locally.
Explore UnicodeFix →
Local secrets management
Secrets Kit
Store credentials locally and pass selected secrets to the process that needs them, reducing the need to keep credentials in project files and startup scripts.
Explore Secrets Kit →More of the toolkit
- Extract Chat — Turn conversation exports into readable Markdown or HTML.
- WebReader for Safari — Listen to web articles through your own text-to-speech setup.
- LLM Ops Kit — Operate and troubleshoot self-hosted AI services.
When the pieces need to work together
A useful tool is only part of a working system. Distributed Thinking Systems also provides hands-on integration, automation, and troubleshooting for local AI and sensitive workflows.
Local AI & agents
Model runtimes, agent workflows, and document processing fitted to your hardware and the data you need to handle.
Secure automation
Shell tools and service workflows that make credential handling explicit and reduce unnecessary copies of sensitive data.
Systems engineering
Unix, Linux, and macOS integration, diagnostics, and developer infrastructure—with attention to how it will be operated and maintained.
Follow the engineering
The writing covers implementation, experiments, failures, and the decisions behind the tools—not just release announcements. Read a complete series, try a Hands-On tutorial, or start with a recent post.
Research with a human purpose
Some of this work grew from the difficulty of organizing sensitive communications and family-law records. Case Analytics and parental alienation research remain part of that longer-term effort. The aim is to support careful review and preserve context, not to turn an AI output into a finding about a family.
For research participation or project contributions, see community collaboration. For background on the company and its work, visit About Us.