Research
We don't claim it. We measure it.
Every paper here tests one promise Penweir makes. We publish the method, the numbers and the limits, so you can check our work instead of trusting it.
How we publish
Three rules for every paper.
One promise per paper
Each report tests a single claim, such as that nothing sensitive leaves the network, and says what would count as failure.
Limits in the open
What the test did not cover is part of the paper, not a footnote. Where the test found a weakness, we say what we are changing.
No content, ever
Test data is synthetic. Reports hold counts, timings and status codes, never prompts or answers.