Projects

The 32 public projects run by our three researchers. 5 personal projects are left off this list. The bars under each card count research journals per week since March 2026.

Active 27On hold 3Done 2As of Oct 9, 2026
AI agents & memoryActive

neuromem

A long-term memory engine that lets an AI agent keep who it is and what it has been through, and recall the memories that fit the moment by association.

MarWeekly journalsOct
Journals42Cards48Last2026.09.26
AI agents & memoryActive

mnemosure

A memory layer that cuts forgotten context and hallucination in AI coding across sessions. It says "I don't know" when it doesn't, and answers with sources when it does.

MarWeekly journalsOct
Journals14Cards30Last2026.09.07
AI agents & memoryActive

numen

A multi-agent runner that launches Claude Code, Codex and Gemini CLI unmodified as boxes on a canvas, wires them together, and shows who started when and what failed.

MarWeekly journalsOct
Journals15Cards22Last2026.08.09
AI agents & memoryActive

honcho-selfhost

Runs the open-source memory server Honcho on our own machines as a long-term memory store shared by Claude, Codex and Gemini.

MarWeekly journalsOct
Journals42Cards9Last2026.10.07
AI agents & memoryActive

Human-memory-based memory

Gathers memory research from cognitive psychology and neuroscience to ground the design of a memory system modeled on human memory.

MarWeekly journalsOct
Journals1Cards6Last2026.07.01
AI agents & memoryActive

Memory systems integration

Draws the boundaries between our five memory projects and weighs on one scale what to keep and what to fold.

MarWeekly journalsOct
Journals0Cards1Last—
LLM evaluationActive

Serialization formats and token accuracy

Feeds the same data to LLMs in seven formats, including JSON, YAML, CSV and Markdown tables, and traces why accuracy and token counts change.

MarWeekly journalsOct
Journals9Cards11Last2026.08.21
LLM evaluationDone

ai-coding-arch

Tested "does architecture help AI coding?" across 15 code structures and 209 cells. What separated cascading failures was not the structure but the model tier.

MarWeekly journalsOct
Journals3Cards6Last2026.09.12
LLM evaluationOn hold

llm-bench

A benchmark that measures on several axes which model fits our own everyday use, rather than public leaderboards.

MarWeekly journalsOct
Journals21Cards28Last2026.08.12
Developer tools & automationActive

agrune

A browser-control layer with no LLM of its own. Any agent connects over MCP and handles web pages by meaning instead of screen coordinates.

MarWeekly journalsOct
Journals21Cards25Last2026.08.16
Developer tools & automationActive

dsforge

Verifies by machine measurement, not human judgment, whether the path from Figma mockups to a design system and code stays faithful.

MarWeekly journalsOct
Journals11Cards14Last2026.09.11
Developer tools & automationActive

dsmonitor

A CLI that scores the consistency of code, styles and design. Published on npm as dsmonitor.

MarWeekly journalsOct
Journals6Cards11Last2026.09.14
Developer tools & automationActive

pidgin

A canvas where AI and people edit meaning together, not coordinates.

MarWeekly journalsOct
Journals5Cards3Last2026.09.12
Developer tools & automationActive

n8n automation

Runs image-generation workflows from Discord commands with n8n alone, without an always-on bot.

MarWeekly journalsOct
Journals1Cards4Last2026.07.21
Developer tools & automationActive

agent-bar

A macOS menu bar tool that shows usage across AI providers at a glance, and draws partial data without distorting what it means.

MarWeekly journalsOct
Journals10Cards1Last2026.10.02
Developer tools & automationActive

Research journal system

This lab's record system: researchers log their work as journals and refine conclusions into a knowledge graph. The numbers on this site come from it.

MarWeekly journalsOct
Journals15Cards39Last2026.09.16
Developer tools & automationActive

Research record framework

Surveyed how universities, companies and institutes keep research records, and set the format of our lab notes and weekly reports.

MarWeekly journalsOct
Journals2Cards8Last2026.07.09
Things we buildActive

3d-fabrication

Builds an isolated Blender environment where AI edits and checks 3D-printing models (STL) directly, then redesigns real prints with it.

MarWeekly journalsOct
Journals3Cards1Last2026.09.11
Things we buildActive

toython

A handheld voice input device with six buttons and a microphone. Transcription runs on local Whisper on a Mac.

MarWeekly journalsOct
Journals3Cards6Last2026.08.27
Things we buildActive

ride-draft

A local web tool for refining bike route drafts on a map and exporting them as GPX for a bike computer.

MarWeekly journalsOct
Journals15Cards18Last2026.08.14
Things we buildActive

wisp

A macOS app that transcribes speech on the device, without sending audio to the cloud.

MarWeekly journalsOct
Journals11Cards9Last2026.08.05
Things we buildActive

aria

A local composing environment where AI writes and plays music, and people listen and edit in the browser.

MarWeekly journalsOct
Journals10Cards4Last2026.09.23
Things we buildActive

locogic

Checks in Godot whether a deterministic puzzle game about opening blocked rail networks with the fewest works holds up.

MarWeekly journalsOct
Journals19Cards16Last2026.07.29
Things we buildActive

sulnote

A web app for whisky, wine and sake tasting notes and prices by shop, with taste analysis.

MarWeekly journalsOct
Journals8Cards7Last2026.10.06
Things we buildActive

novel

An experiment in writing a Korean web novel all the way to the end with AI.

MarWeekly journalsOct
Journals9Cards9Last2026.06.25
Things we buildActive

broker-rogue

A roguelite stock-survival game where you move the market with daily trading cards. Built with React Native and Expo.

MarWeekly journalsOct
Journals1Cards0Last2026.09.14
Things we buildActive

ktheme-works

Generates KakaoTalk themes from code instead of drawing them by hand. Change one color and everything regenerates.

MarWeekly journalsOct
Journals4Cards14Last2026.09.16
Things we buildActive

Image generation pipeline

Builds SDXL and ComfyUI image-generation presets that give the same result no matter who runs them.

MarWeekly journalsOct
Journals12Cards10Last2026.10.07
Things we buildOn hold

screen-translate

A Chrome extension and Mac app that turns foreign text on screen into Korean in place.

MarWeekly journalsOct
Journals7Cards10Last2026.06.25
Things we buildOn hold

jarvis

A local voice assistant that adds only voice and a HUD on top of the Codex we already use.

MarWeekly journalsOct
Journals3Cards4Last2026.06.25
Learning togetherActive

rag-chatbot-study

A study project building a RAG chatbot that answers from uploaded documents.

MarWeekly journalsOct
Journals1Cards0Last2026.09.14
Learning togetherDone

ax-hackathon

Joined the "AX Talent War" hackathon by JocodingAX and Primer, and built Codex plugins that solve each company's problem.

MarWeekly journalsOct
Journals12Cards13Last2026.07.23

No projects match these filters.

Numbers are counted from the research journal repository. Journals is the number of research journals written for a project; Cards is the number of finding, hypothesis and decision cards drawn from it.

Curious about a project?

Questions about our research, collaboration and requests for materials are all welcome by email.