Agent tools, memory systems, documentation pipelines, prediction-market research, MUD tooling, and small weird interfaces.
I build tools for AI-assisted development: better context, better coordination, richer memory, and sharper command surfaces for agents.
| Project | What it does |
|---|---|
| html-home | Gives repo-local HTML artifacts one localhost front door through manifests, a local index, and a read-only server. |
| pro-cli | ChatGPT Pro and Deep Research in your terminal. |
| raindrop-cli | Exposes Raindrop bookmarks through a clean agent-friendly CLI. |
| discive | Downloads entire Discord channels for local archives and analysis. |
| extra-eyes | Spins up N pair programmers with custom profiles to give instant feedback to one context window. |
| liquid-mail | Shared mailboxes for agents: progress logs, decisions, and coordination context that survive across sessions. |
| knb | Builds knowledge-base workflows for research, synthesis, and durable context. |
| salt-ai | A tiny utility for salting LLM responses. |
| annas-archive-ebooks | Adds ebook search and download workflows for agent-assisted research. |
| claude-skills | A collection of some of my favorite agent skills. |
| ascii-love | Plays with terminal-native visual interfaces and ASCII display work. |
Villagecraft was a community calendar and organizing platform for neighborhood-scale workshops, skillshares, and hands-on projects.
The work centered on ground-up organizing: helping people coordinate groups, share resources, find spaces, spread the word, and make local co-learning visible. The old FAQ framed it as tools for "hands-on learning activities at the neighborhood level" and a network for co-creation, creative projects, wellness workshops, and community-led making.
| Thread | Repos |
|---|---|
| Agent context | slurp-ai, claude-skills, annas-archive-ebooks |
| Agent command surfaces | pro-cli, raindrop-cli, discive |
| Agent coordination | liquid-mail, pair-programmer-ai, openclaw-hindsight |
| Agent experiments | extra-eyes, knb, ai-isms, dianalokada-skill, gloss, rivet, salt-ai |
| Thread | Repos |
|---|---|
| Browser and visual context | glimpsy, ascii-love, itoa |
| GitHub display experiments | github-marquee |
| Text and social archives | jeffrey-emanuels-tweets |
| Thread | Repos |
|---|---|
| Ranvier and MUD ecosystem | mudcolors, mudcolors-tester, ranvier-community, ranvier-awesome, ranvier-multi-arg-parser, area-respawn |
| Enceladus and game work | enceladusgame, enceladus-client, evermore-color-staging |
| Older web games and exercises | catchphrase, ToEat.ly, bitly_clone, burger_app, json_blob_challenge |
| Thread | Repos |
|---|---|
| Agent and AI systems | ypi, codex-orchestrator, stealth-browser-mcp, LlamaIndexTS, tokens, exa-cli |
| Prediction markets and crypto | pmxt, testnets |
| MUD/Ranvier ecosystem | ranviermud, pinwheel, core, awesome-ranvier, axolemma, neuro, dome-client.js, lobus |
Named without links:
| Area | Projects |
|---|---|
| Prediction-market research and trading systems | pdb, pm, pm-weather, polymarket, trading, econ, pmknb, research |
| Agent memory and knowledge systems | dev-decision-traces, knowledge-scanner, ai-deas, magic-treehouse, tremors |
| MUD automation and game agents | mudbot, mudbot-btree, Our-Mortal-Haul, rune-age-sagas, eterna, ember |
| Browser, media, and research tools | clearfeed, clearfeed-extension, replyhunter, media-scanner, text-crawler, twitter |
| Crypto and finance experiments | sol-fleet, crypto-hourly, xcapit-dist, cryptonation, uniswap-trader |
| Older worldbuilding and web projects | enceladus, enceladus-site, evermore-web, evermore-backup, mudlense, frontman, frontm, puntman, crowd, crowd-symbols, ketascrape |