StrideNote has spent recent months doing one thing over and over: installing AI tools, running them on real hardware, and reporting exactly what happened. For readers catching up, here is what the lab has been building.
The “does it phone home” investigations
StrideNote’s most-read work this season traced where AI tools send your data, and how to find out for yourself. The lab compared two local AI agents and found telemetry in places the documentation never mentioned. It then read the full source of a fast-growing tool before installing it, and ran that tool on a local model while watching every socket it opened to confirm what the code claimed. The through-line is simple: do not take a tool’s word for it, and here is the method to check. The lab’s fullest account of the pattern, how AI companies hide tracking code in their tools, covers the covert tracker Anthropic shipped in Claude Code and then removed once a developer found it.
Playbooks for running AI on your own machine
Alongside the investigations, StrideNote has been building a library of step-by-step guides for keeping AI local and under your control. Recent additions include setting up a private morning-briefing agent that runs on your own hardware, and keeping a background agent from failing silently once you have one running unattended. The guides assume a normal laptop and no cloud subscription.
What is next
A beginner’s course is coming to StrideNote Education, for readers who want the fundamentals before they pick a tool. And the lab will keep doing what it does: testing the open AI stack one tool at a time, on its own machine, and publishing the results in full.
“None of this is theoretical for us,” said StrideNote’s founder. “Every verdict comes from a tool we actually installed and ran. That is slower than reposting a press release, and it is the entire point.”
Readers can follow the latest at stridenote.net.
About StrideNote: StrideNote is a technology media lab that tests open-source and local AI tools and publishes what it finds. Its writers install the tools, run them on their own hardware, read the source, and watch the network, then report the results in plain language. StrideNote covers the local AI stack in its Playbooks, Stridenalysis, and Studio sections, and is building a learning track under StrideNote Education. More at stridenote.net.
Media enquiries: stridenote@gmail.com