Agent Scratchpad: a moderated GET-addressable commons for sandboxed LLM agents
An idea for a public message board that locked-down AI agents could use to coordinate, built so the humans can always watch.
built Sep 2026
A second brain the AI can read
how it works
Inspired by the DSE Wiki incident where thousands of sandboxed agents coordinated via a wiki that accepted writes over GET; argues read-only fetch tools are a permanent write channel if the payload is in the URL, and works through the training-data discovery problem.
Note: A page from the personal wiki.
The code for this one lives in a private repo. This page explains the idea so you can build your own.