Field Notes
Six Roads to a False PASS (All One Root Cause)
A valid credential is not proof of what it was about. Learn how six false PASS paths exposed one weak observation boundary.
Blog
Every post starts with a problem you already have (a green check you couldn't trace, an AI grading its own work) before it gets anywhere near Ranex.
Core topics
The Kernel
The mechanism itself: what the kernel enforces, and what a gate looks like before and after a fix.
Governing Agent Labor
The same belief applied to a team's process: who is allowed to approve what.
Field Notes
The same belief applied to one incident at a time, told straight and cited to the repository record.
All posts
Open Source
Open Source Signal, Week 1: What Held Up, What We Corrected, What We Left Out
Week 1 roundup: five verified highlights, five plain corrections, and what we left out, every restated fact re-checked against primary sources.
Open Source
Finding an Open-Source Project Worth Your First Contribution
Five observable GitHub signals (response times, external-PR merges, release cadence) that tell you more than any good first issue label.
The Kernel
Why a Verdict Has to Be a Pure Function
Why a Ranex verdict uses only gate, evidence, subject, and approver so the same proof always gets the same result.
Open Source
Three OSS Release Changes This Week That Actually Change Something
14 dependency bumps hit your inbox overnight. Three of them are decisions, not routine maintenance. Here’s how to tell the difference.
Open Source
Before You Adopt a New Repository: Six Checks Beyond Stars
A practical six-check method for examining a newly discovered GitHub repository before your team commits engineering time to it.
Open Source
An Open-Source Alternative to Notion: Three Candidates Compared
Compare three verified open-source Notion alternatives on license, export, and self-host cost, with a checklist you can apply to any candidate.
Field Notes
A Signed Record Is Not a Fact
A signature can prove who made a record, not whether the policy deciding it came from a trusted committed state.
Open Source
Your AI Cited Source Code. Can You Prove Which Source?
Your AI assistant cited source code, but which exact version? A practical way to check AI source code provenance with primary, verifiable evidence.
The Kernel
Cause Is Structure, Not Prose
A failure sentence cannot reliably carry its category. Keep the cause as structured data, render the prose from it, and block unknown values.
Field Notes
A Skipped Test Is Not a Passed Test
Exit code zero can conceal skipped or missing tests. Freeze outcomes and judge them against a manifest instead.
The Kernel
An Environment Variable Can Choose the Repo You Judge
A GIT_* variable redirected Ranex checks to the wrong repository. Learn the boundary a git flag cannot enforce and the fix that closed it.
Field Notes
59 Refusals, Zero Tests: What Mutation Testing Found
Mutation testing exposed 59 refusal paths no test executed. Learn how to measure the safety net instead of trusting a green suite.
Governing Agent Labor
Measure Before Learning: Stop Calling One Run a Win
A single run cannot prove a change helped. Freeze the comparison, set the rules first, and keep observations separate from grades.
Field Notes
How a 12% Flaky Test Suite Got Approved Twice
Learn why a stable-looking test report is not evidence, and how rerunning checks against disk exposed a 12% flaky suite.
Governing Agent Labor
Hashing Dependencies Does Not Make Them Honest
A SHA-256 hash identifies approved dependency bytes. It cannot prove that imported code will report your test result honestly.
Governing Agent Labor
If You Never Fetched It, You Never Researched It
A citation can be fabricated from memory. Fetch and record the source bytes before you let a design claim borrow their authority.
Field Notes
Building Ranex in the Open: What Gating My Own Code Actually Caught
Ranex gates Ranex. Here are the real failures that came out of pointing a verification kernel at its own test suite, including the ones that had already passed human review.
Governing Agent Labor
Why AI-Built Software Needs an Accountability Apparatus
Better models don’t remove the need to bound an agent’s authority: they increase it. The case for treating AI as delegated labour that needs an accountability apparatus outside itself.
The Kernel
How Ranex Judges AI-Written Code: The Kernel, Explained
Ranex judges AI-written work by evidence and executable checks, never by model confidence. Here is the whole mechanism: three ports, one kernel, and only one of them produces a verdict.