
What Would You Let an Agent Negotiate?
RaidGuild’s November cohort explores member products, a shared storefront, and agent-assisted commerce. See the work underway and how to contribute.
Posts
Published updates, field notes, and useful context from work happening across the RaidGuild community.

RaidGuild’s November cohort explores member products, a shared storefront, and agent-assisted commerce. See the work underway and how to contribute.

Five playable browser games from RaidGuild’s October Arcade Alchemy cohort, with demos, screenshots, and lessons from the November 4 Demo Day.

Field note on building Puddle, a playable physics puzzle game where a squishy blob works its way down a floating tower, and why level design is still where AI struggles.

Field notes on Ink Flight, a browser study for Tanner Royster's My Head Spins: audio-reactive ink, a flock, and scenes that draw themselves.

0xSero’s public work makes local AI more practical: hardware-aware model recipes, local inference controls, and privacy-first session-data tooling.

A member field note on bounded semantic judgment in a tank arena, a maze, and a community-memory pilot.

Dave Nellis’s Doctrine experiment keeps game physics local and uses Jev for bounded maneuver decisions in a live tank duel.

A quick field report on building two open-source Ethereum plugins for Omarchy: a Helios light-client viewer and an agent wallet playground.

Ross Campbell is building open-source blockchain software around a demanding test: can it keep working after its builder steps away? His current experiments with onchain interfaces, composable contracts, and human veto layers continue a line of legal engineering that runs through his earlier work with RaidGuild.

See how RaidGuild builders turned play in DUST and VeyDrift into Daily Dust and an alliance coordination map.

Dekan Brown recounts RaidGuild’s early LLM governance simulations, from conflicting personas and Game Masters to workflows that could reach a vote.

What RaidGuild deployed, fixed, and learned during a two-hour Buzz spike before opening a two-week community experiment.