Writing

What I built, what broke, and the numbers. New pieces also go out by RSS.

  1. 3,380 consecutive provider failures were a policy failure, not a transient one

    My model router kept sending traffic to a provider with no credit for 3,380 failures in a row, because a bandit treats failure as a score and not as a reason to stop.

    · 6 min read

  2. 80 Access apps and one wildcard

    A Cloudflare Worker that provisions Zero Trust Access apps from a JSON config, never touches group membership, and audits the whole account for broad policies.

    · 7 min read

  3. A commit gate you can watch fail

    Three commit-time gates (secrets, private terms, tokens in git URLs) published with the evidence that each one blocks bad input, including the false positive that read SVG coordinates as phone numbers.

    · 8 min read

  4. A Durable Object per GPU node, for a home Mac mini

    How a Cloudflare Worker queues, preempts and recovers jobs for a home GPU box that sleeps, drops its tunnel and restarts whenever it likes.

    · 7 min read

  5. A merge queue that only fast-forwards trunk to a commit it verified

    I rebuilt a small land queue so the commit trunk advances to is, by construction, the commit the checks ran on, and gave it a shadow mode to run before you trust it.

    · 7 min read

  6. A red team that has to quote its sources

    A layered red team over policy documents where every quote is checked against the source text, and one unsourced quote makes the run exit 1.

    · 6 min read

  7. A test that refuses to ship an LLM stage nobody measured

    I wired a stage evaluation into cargo test, so a registered LLM stage with no measured verdict, or a prompt edited since its measurement, fails the build.

    · 7 min read

  8. Admission by reservation, not instantaneous load

    A scheduler that checked current load let four jobs onto a Mac that then hit load 44, so I made every job reserve its peak before it starts.

    · 7 min read

  9. An on-device LLM as autocorrect, with a typed did-you-mean

    I pointed Apple's on-device 3B model at autocorrect, forced its answer into a typed schema, and found out how little of that I've actually measured.

    · 6 min read

  10. Caching made my rewrite pipeline cost more, and the cheap-model verdict only half held

    A three-judge panel scored 810 rewrite candidates; caching raised cost at no quality gain, and a fallback backend I hadn't noticed was skewing the model comparison.

    · 7 min read

  11. Cost in the reward, once

    armwright learns which model to send each kind of task to, from the outcomes of real calls. What a seeded 20,000-call simulation shows about where the cost knob really is.

    · 3 min read

  12. How goldenwatch refuses a release

    Most eval scripts fail open. goldenwatch names six ways a golden-set check goes quietly wrong, and refuses each one.

    · 3 min read

  13. Mutation testing scoped to a diff, and why survivors need a written reason

    A mutation-testing tool that mutates only the functions a diff touched, runs every mutant in a throwaway sandbox, and refuses to ignore a survivor without a written reason.

    · 7 min read

  14. My AbortSignal was correct and the call still hung for six hours

    A model call hung twice for about six hours behind a correct AbortController, so I moved the deadline to a timer that does not depend on the signal.

    · 6 min read

  15. Patching a ReDoS in picomatch without forking the world

    A glob library under most JavaScript build tools had an unfixed regex advisory, so I refused one construct, kept every other byte identical, and made the copy fail the build if it drifts.

    · 6 min read

  16. Porting a Python gate to Go with golden files, and why append-only ledgers merge cleanly

    I replaced a Python checker for a decisions graph with a stdlib-only Go tool, and checked it against the old output byte for byte using recorded files.

    · 6 min read

  17. Redact before the model, restore after

    A reversible PII boundary around every model call in a chat app, tested by planting identifiers in the input and failing on any that show up in what leaves.

    · 6 min read

  18. Six guards in one day that had never been shown to fail

    A log of six checks in my own repos that looked present and had never been tested against bad input, and the one habit that exposed them.

    · 7 min read

  19. Temperature silently does nothing on reasoning models, and the API will not tell you

    A per-provider matrix of what temperature does while a model is thinking, and a small Rust library that sends, drops or refuses it and records which.

    · 7 min read

  20. Testing firmware logic on a laptop with a port

    I put the timing, retry and sequence logic of a two-board hobby project behind a three-method interface, so most of it runs and breaks on my laptop instead of on a board.

    · 6 min read

  21. Three P0s that looked wired

    An alert sender with no credentials, an auth check that returned true, and a secret rotator that returned constants, all found in one audit.

    · 6 min read