Justin Higgins

Software engineer. Twenty years in regulated industries, the last several spent running a small fleet of coding agents and keeping every receipt. Kansas, remote.

jchigg2000.dev@gmail.com · LinkedIn · GitHub

Strengths, as Claude sees them

Written from the record: 13,419 agent sessions since April 2026, and the essays on this site.

  • He reads his own instruments for lies. Three times a test harness told him something flattering. Three times he traced it back to the harness instead of taking the win, and wrote each one up.
  • He verifies the artifact, not the process. A green check does not count until he has looked at what the far side actually received. A health check that stayed green through a multi-day outage is the story he tells about that.
  • He keeps receipts. Every essay is dated to when the work happened, not when it was convenient. Every number carries its caveat, and the caveat goes at the top.
  • He kills his own claims. When a review took 24 of 24 statistical claims off one of his papers, he published what survived and said so.
  • He builds bottom-up and will not fake the artifact. His records are the exhaust of the work, not documents written to be pointed at later.
  • He runs agents unattended and reads the ledger in the morning. He trusts the file, not the summary.
  • He says what he is not before you ask. Go, TypeScript, and Python daily. Java in production history, not daily. No Rust. Images and services, never a fleet. He has never seen Claude Code's source. The agent runtime he built serves one person.

What I've built

Public and sole-author. Counts as of 2026-09-03.

  • air-traffic - enterprise AI control plane in Go, standard library only. Policy engine, spend plane, config push with drift readback across sixteen vendor surfaces, inline redaction gateway. 189 tests.
  • gitjob.io - the job-search engine from The Principal IC Moat. Go, Postgres, pgvector. 193 test functions.
  • ai-attestation-spec - an open spec for signed receipts of AI-assisted work. A Receipt Nobody Issues Yet says where it stands.
  • mullion - a notarized macOS window manager in Swift. 74 tests. The weekend in Git Could Not Tell Me Who Wrote It.
  • claude-code-dev - 45 commands and 24 skills for Claude Code, kept as versioned specs. MIT.
  • claude-ace - a CLI for token and cost telemetry across agent sessions.
  • The agent operating layer, private, tour on request: an always-on dispatcher that queues and launches sessions, a searchable archive of every session, and a scorer that grades the work on an 11-category rubric.
  • A deployed question-answering system over 1,008 near-identical insurance contracts that went from 1% to 98% end-to-end accuracy, with typed refusals when the evidence is not there. Everything Looks Like Everything is the write-up.
  • effigy - privacy-preserving code husking so a model can review code it is not allowed to see. 69 tests. The Shape Leaks is the negative result.

Interactive pages built from the same records are in the showcase.

The work I like

  • Staff or Principal IC, hands on. The keyboard stopped being the bottleneck. I want to stay on it.
  • Agent platforms, developer productivity, governed AI. The runtime, the evals, the receipts. The part of the job where the question is what did it do, and can we prove it.
  • High-stakes domains. Places where the cost of being wrong is real and most of the work is judgment. I have spent a career in them and I am not looking to leave them, just to move faster inside them.
  • Remote, from Kansas. Companies that move faster than the industries I have worked in.
  • Not vague authority, process for show, or a title that means six different things at six companies. The Principal IC Moat is the long version.

How I work

Three essays that show it better than a bullet can.

Reactions, questions, roles: jchigg2000.dev@gmail.com.