https://ranex.dev/blog Deterministic governance for AI agents that build software. Ranex judges work by evidence and executable checks, never by model confidence. en Wed, 26 Aug 2026 09:00:00 GMT https://ranex.dev/blog/environment-boundary-for-git-queries https://ranex.dev/blog/environment-boundary-for-git-queries A GIT_* variable redirected Ranex checks to the wrong repository. Learn the boundary a git flag cannot enforce and the fix that closed it. Wed, 26 Aug 2026 09:00:00 GMT Anthony Garces The Kernel Evidence Supply Chain https://ranex.dev/blog/59-refusals-no-test https://ranex.dev/blog/59-refusals-no-test Mutation testing exposed 59 refusal paths no test executed. Learn how to measure the safety net instead of trusting a green suite. Mon, 24 Aug 2026 09:00:00 GMT Anthony Garces Field Notes Agent Safety Mutation Testing Test Design https://ranex.dev/blog/measure-before-learning https://ranex.dev/blog/measure-before-learning A single run cannot prove a change helped. Freeze the comparison, set the rules first, and keep observations separate from grades. Fri, 21 Aug 2026 09:00:00 GMT Anthony Garces Governing Agent Labor Benchmarks Evidence Test Design https://ranex.dev/blog/flaky-suite-approved-twice https://ranex.dev/blog/flaky-suite-approved-twice Learn why a stable-looking test report is not evidence, and how rerunning checks against disk exposed a 12% flaky suite. Wed, 19 Aug 2026 09:00:00 GMT Anthony Garces Field Notes Evidence Flaky Tests Test Design https://ranex.dev/blog/hashing-dependencies-doesnt-make-them-honest https://ranex.dev/blog/hashing-dependencies-doesnt-make-them-honest A SHA-256 hash identifies approved dependency bytes. It cannot prove that imported code will report your test result honestly. Mon, 17 Aug 2026 09:00:00 GMT Anthony Garces Governing Agent Labor Evidence Hashing Supply Chain https://ranex.dev/blog/never-fetched-it-never-researched-it https://ranex.dev/blog/never-fetched-it-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. Fri, 14 Aug 2026 09:00:00 GMT Anthony Garces Governing Agent Labor Agent Safety Evidence https://ranex.dev/blog/building-ranex-slice-log https://ranex.dev/blog/building-ranex-slice-log 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. Thu, 13 Aug 2026 13:01:57 GMT Anthony Garces Field Notes Evidence Gate Verdicts Slice Log https://ranex.dev/blog/ai-accountability-apparatus https://ranex.dev/blog/ai-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. Thu, 13 Aug 2026 13:01:54 GMT Anthony Garces Governing Agent Labor Agent Safety AI Accountability Separation of Duties https://ranex.dev/blog/how-the-kernel-works https://ranex.dev/blog/how-the-kernel-works 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. Thu, 13 Aug 2026 13:01:48 GMT Anthony Garces The Kernel AI Accountability Evidence Gate Verdicts