Eryk Orłowski
CV · ContactCV and contact

Eryk Orłowski · hretheumhretheum.com/questions/built-with-agents

Question P6 of 8

All questions6

What have they built with agents on their own?

Comes up with, designs and builds their own tools with AI agents, alone. Kelter, an engine for running agents, has 291 tests and 32 recorded architecture decisions, and does the same work on a local 0.5B model and on a frontier model after changing 1 line of configuration. Gatefold, a design process with AI in which a person makes every decision, was built in 131 commits between 4 and 13 October 2026, and 12 patterns accepted in it became a course on this site's subdomains. The engine of this site gives only facts from the claim package; in a measurement on a real model, the chat gave no fact from outside the package.source sources: om-05, ke-01, gf-01, fo-01, ke-04, ke-03, gf-04, gf-05, fo-02, fo-04

Evidence

  1. Kelter has 291 tests and 32 recorded architectural decisions.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: ke-04case study →case study: hretheum.com/work/kelter

  2. Measured on the real model, the chat gave no facts from outside the package, fell back in 3.3% of answers and took 9.5 s for the slowest 5% of answers.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 10.2026sourcesource: fo-04case study →case study: hretheum.com/work/folio-engine

  3. It depends on what hands-on means to you. […]It depends on what hands-on means to you. For me today it means building. I build my own tools with AI agents: Kelter, an engine for running agents; Gromatik, a platform for maintaining design systems; and a set of twelve tools that measures brand consistency for a client with many brands. I came up with them, designed them and built them myself. So I'm not someone who stopped making things and now only draws org charts. The kind of making has changed, and I'd argue it's the kind a team going into AI needs most.”

    answersourcesource: ko-n7-1

  4. 12 patterns accepted in Gatefold became the course on this site's job-posting subdomains, in production since 10 October 2026.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 10.2026sourcesource: gf-05case study →case study: hretheum.com/work/gatefold

  5. On this site every fact comes from a profile package of claims with an identifier, a word-for-word quote and a status; only confirmed claims go in.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 10.2026sourcesource: fo-02case study →case study: hretheum.com/work/folio-engine

  6. Gromatik was built in 256 commits between 5 August and 13 September 2026, with about 1000 passing tests.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: gr-05case study →case study: hretheum.com/work/gromatik

  7. I build tools with AI agents myself: Kelter, Gromatik and Exocortex.”

    profilesourcesource: om-05

  8. I know CI/CD principles from practice: in my own projects I run pipelines across dozens of repositories: tests and image builds, live contract checks, releases that can be rolled back, automated dependency updates and fork syncing.”

    profilesourcesource: om-08

  9. The toolchain is a monorepo of 12 command-line tools with 78 test files, built in June 2026.”

    projekt, sprawdzone w kodzie · me · to be confirmedEnergy group with many brands, 2026sourcesource: ds-03case study →case study: hretheum.com/work/multi-brand-design-system

  10. I came up with, designed and built the toolchain myself, working with AI agents.”

    projekt · me · to be confirmedEnergy group with many brands, 2026sourcesource: ds-07case study →case study: hretheum.com/work/multi-brand-design-system

  11. I came up with, designed and built Kelter myself, with AI agents; it is my intellectual property.”

    własne narzędzie · me · to be confirmedOwn project, 2026sourcesource: ke-01case study →case study: hretheum.com/work/kelter

  12. In Kelter the kernel is the runtime and the language model is only a policy.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: ke-02case study →case study: hretheum.com/work/kelter

  13. The same job runs on a 0.5 B local model and on a frontier model after a change of 1 line of configuration.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: ke-03case study →case study: hretheum.com/work/kelter

  14. Kelter will be used in Exocortex R&D and eventually opened to the public.”

    własne narzędzie · me · to be confirmedOwn project, 2026sourcesource: ke-05case study →case study: hretheum.com/work/kelter

  15. I designed and built Gromatik and blind.ds myself, as a private project.”

    własne narzędzie · me · to be confirmedOwn project, 2026sourcesource: gr-01case study →case study: hretheum.com/work/gromatik

  16. Generating component code is one model call per component, with no agent loop and no code execution on the server.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: gr-02case study →case study: hretheum.com/work/gromatik

  17. Every run records its settings and a fingerprint of its input, so you can trace where a result came from.”

    własne narzędzie · me · to be confirmedOwn project, 2026sourcesource: gr-06case study →case study: hretheum.com/work/gromatik

  18. Exocortex pre-synthesises threads from different sources around a client, a project, a professional domain or a field of knowledge.”

    własne narzędzie · me · to be confirmedOwn project, 2026sourcesource: ex-01case study →case study: hretheum.com/work/exocortex

  19. Several dozen documents a day go through Exocortex in total.”

    własne narzędzie · me · to be confirmedOwn project, 2026sourcesource: ex-03case study →case study: hretheum.com/work/exocortex

  20. The lab is a separate, public project that uses only the Exocortex engine and is cut off from confidential and private data.”

    własne narzędzie · me · to be confirmedOwn project, 2026sourcesource: ex-04case study →case study: hretheum.com/work/exocortex

  21. The knowledge graph runs on PostgreSQL with the Apache AGE extension.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: ex-05case study →case study: hretheum.com/work/exocortex

  22. I designed and built Exocortex myself.”

    własne narzędzie · me · to be confirmedOwn project, 2026sourcesource: ex-06case study →case study: hretheum.com/work/exocortex

  23. The hypothesis card is frozen and its checksum published before anything is calculated.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 09.2026sourcesource: rd-01case study →case study: hretheum.com/work/exocortex-rnd

  24. No file reaches the public repository without 5 automated checks of the publication gate.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 09.2026sourcesource: rd-02case study →case study: hretheum.com/work/exocortex-rnd

  25. The publication gate is tested every night on files with deliberately planted leaks.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 09.2026sourcesource: rd-03case study →case study: hretheum.com/work/exocortex-rnd

  26. The lab's isolation from the private instance is checked every night by a test from inside the lab.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 09.2026sourcesource: rd-04case study →case study: hretheum.com/work/exocortex-rnd

  27. In the first full week 198 papers went through the opportunity radar.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 09.2026sourcesource: rd-05case study →case study: hretheum.com/work/exocortex-rnd

  28. I am the project owner and run the execution through agents.”

    własne narzędzie · me · to be confirmedOwn project, from 09.2026sourcesource: rd-07case study →case study: hretheum.com/work/exocortex-rnd

  29. I designed and built Halvern, a fork of the open-source Meetily, working with AI agents.”

    własne narzędzie · me · to be confirmedOwn project, 2026sourcesource: hv-01case study →case study: hretheum.com/work/halvern

  30. Halvern records, transcribes and summarises meetings entirely on the user's computer, with no account and no upload.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: hv-02case study →case study: hretheum.com/work/halvern

  31. In Halvern onboarding asks which language meetings are in, and the answer picks the transcription engine.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: hv-03case study →case study: hretheum.com/work/halvern

  32. Halvern checks for updates once at start-up, not on a timer, so it does not reveal when the user is in meetings.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: hv-04case study →case study: hretheum.com/work/halvern

  33. 179 of 310 commits of Halvern were co-written with Claude.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: hv-05case study →case study: hretheum.com/work/halvern

  34. Halvern is a signed and notarised pre-release for Macs with Apple Silicon, with a public repository under the MIT licence.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: hv-07case study →case study: hretheum.com/work/halvern

  35. I came up with, designed and built Gatefold myself, working with AI agents.”

    własne narzędzie · me · to be confirmedOwn project, from 10.2026sourcesource: gf-01case study →case study: hretheum.com/work/gatefold

  36. Gatefold was built in 131 commits on the main branch between 4 and 13 October 2026, with 32 recorded architectural decisions and about 500 tests.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 10.2026sourcesource: gf-04case study →case study: hretheum.com/work/gatefold

  37. Designing 3 motion sheets with an agent cost 0.80 USD, passing 8 ready scripts through the agent cost 3.34 USD, and a plain program now does that step for 0 USD.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 10.2026sourcesource: gf-07case study →case study: hretheum.com/work/gatefold

  38. I came up with, designed and built the engine behind this site myself, working with AI agents.”

    własne narzędzie · me · to be confirmedOwn project, from 10.2026sourcesource: fo-01case study →case study: hretheum.com/work/folio-engine

  39. For a job posting the engine matches each requirement to up to 3 claims and shows requirements without a match openly as gaps.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 10.2026sourcesource: fo-03case study →case study: hretheum.com/work/folio-engine

  40. 216 of 235 commits of the site rebuild since 1 October 2026 were written with Claude.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 10.2026sourcesource: fo-05case study →case study: hretheum.com/work/folio-engine

  41. Before a merge, a gate in CI compares 60 views of the live site to prove that design concepts do not change it.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, from 10.2026sourcesource: fo-06case study →case study: hretheum.com/work/folio-engine

  42. Cognitive Mirror is an AI session that diagnoses how an adult learns and ends with one change to make.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: cm-01case study →case study: hretheum.com/work/cognitive-mirror

  43. In Cognitive Mirror the model only collects observations, and a rule in code works out the score and confidence level.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: cm-02case study →case study: hretheum.com/work/cognitive-mirror

  44. Splitting the analysis into a separate call per dimension raised the small model's agreement from 20 to 25 of 26 points.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: cm-03case study →case study: hretheum.com/work/cognitive-mirror

  45. In the paraphrase test the production model gave the same result in 39 of 40 cases.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: cm-06case study →case study: hretheum.com/work/cognitive-mirror

  46. The Cognitive Mirror interface is designed in stages closed by my decision, in 4 languages, and checked by 7473 automated tests.”

    własne narzędzie, sprawdzone w kodzie · me · to be confirmedOwn project, 2026sourcesource: cm-07case study →case study: hretheum.com/work/cognitive-mirror

  47. I see my own projects as my R&D rather than a second job competing for time. […]I see my own projects as my R&D rather than a second job competing for time. They're where I learn things before I need them at work. Building an agent engine and a design-system platform taught me how AI systems fail in practice, and that's exactly what I bring to a team adopting AI. The habit of setting a success criterion before measuring anything comes from my own lab and now shapes how my team decides whether an AI workflow is worth keeping. So the honest answer is that those projects make me better at this job.”

    answersourcesource: ko-n10-1

  48. CI/CD pipelines are an everyday thing for me, because almost every project of my own has several. […]CI/CD pipelines are an everyday thing for me, because almost every project of my own has several. I run them across dozens of repositories on GitHub Actions, and the approach differs by project. In the engine behind this site every change goes through tests, a visual-parity check and end-to-end tests, then a container image is built, and the deploy remembers the previous version, so a rollback is a single command. In the demo apps for web, iOS and Android each platform has its own pipeline and its own check guarding the main branch. In projects with AI agents I batch changes into queues tested on separate branches, and the main branch moves only after a green result. On top of that come automated dependency updates, releases with installers and fork syncing.”

    answersourcesource: ko-ci1-1

Limit

Cognitive Mirror has had no real users, and its instrument is not empirically validated.”

Own project, 2026sourcesource: cm-08

E-mail: [email protected] (kt-01) · LinkedIn: linkedin.com/in/eofek (kt-02)

Contact · #contact

I work on AI in 3 areas: AI in the design process, a product or tool with AI, and design systems with AI.

sourcekt-03

Three ways · no form
E-maileof@offline.plkt-01
LinkedInlinkedin.com/in/eofekkt-02
CVDownload CV (PDF, 84 KB)

Built on a 12-column grid. Every fact leads to its source; where there is none, the page says so.