Grok memory: why Grok forgets with memory on, and how we use Grok without it
Grok memory explained: what grok.com remembers, why it seems forgetful, how to turn it off, how it differs from ChatGPT and Grok Bot memory, and MCP.
Blog
Notes from building Tokenrip, the shared workspace your AI agents work from.
Muse custom connector: how to add an MCP server to Meta Muse, and where its memory lives
Meta Muse has no MCP settings form: you ask Muse to build a custom connector. How that works, what Meta reviews, and where Muse keeps keys and memory.
ChatGPT dots memory: what your dot remembers, what you can't see, and what we found on the first call
ChatGPT dots memory explained: what your dot keeps from ChatGPT, the notes you can't see or edit, how to reset it, what it shares with Codex, and MCP plugins.
Claude Duet vs dots vs Muse vs Grok Bot (October 2026): the comparison, and which one we actually use
Claude Duet, OpenAI dots, Meta Muse and Grok Bot compared on price, regions, memory, custom MCP servers and team use, and why Grok Bot won for us.
How to add MCP to ChatGPT (2026): Plugins, Developer mode, and which plans get write tools
How to add an MCP server to ChatGPT: the Plugins flow, Developer mode, which plans qualify, OAuth, local servers, and fixes for the usual problems.
How to use Claude and ChatGPT together (2026): what connects, what doesn't, and how we run both
Claude and ChatGPT can't read each other's chats or memory. What works instead, from copy and paste to one shared MCP workspace, and how we split the work.
Claude Code for teams: what git shares, what stays private, and how two founders run it
Claude Code for teams in 2026: what you share through the repo, what stays per person, the Team plan, agent teams mode, and how our two founders run it.
Claude Code handoff: the prompt, a /handoff skill, and how we do it without a file
Claude Code handoff vs /compact, continuing a session on another machine or your phone, a copyable handoff prompt, and how we hand off between agents.
Claude Code memory: how it works, and why we turned auto memory off
How Claude Code memory works: CLAUDE.md vs auto memory, the 200-line limit, where files live, and how we share context across machines and tools.
Codex vs Claude Code (November 2026): Prices, Limits, Verdict
Codex vs Claude Code, checked November 6, 2026: plans, usage limits, the Codex CLI, memory and sandbox defaults, and why we moved all our coding to Claude Code.