Context Publishing
You built a context graph. Now make it queryable by others — their Claude Code, their agents, their workflows. That's context publishing. We're building the infrastructure for it.
A data engineer built his own intelligence system from scratch just to keep up with AI engineering. That's not unusual. The practitioners who stay current aren't reading more — they're building systems that do it for them.
The next step is sharing what you've built. Right now, the tools don't support it properly. You screenshot Claude Code and send it in Slack. You copy-paste your CLAUDE.md into a doc. You re-explain your setup to every new teammate.
Context publishing is the fix. Your knowledge graph as a live queryable endpoint — for you, your team, and eventually anyone whose agents would benefit from what you know.
What context publishing makes possible
Share your Claude Code setup with your team
One context graph. Everyone pulls from it. Agents included. No more inconsistent outputs across team members re-explaining the same context.
Let others' agents query your expertise
Your methodology, your POV, your domain knowledge — live at an MCP endpoint. Queryable at inference time without you being in the loop.
Get paid for programmatic access to what you know
Context is currently sold as content — one-time. Context publishing makes it a service. Subscribers query your live, evolving graph. You get paid ongoing.
Join the early access list
Context publishing is in active development. If you're building with agents and want to be first when it ships, get on the list.
Questions
What is context publishing?
Making your structured knowledge graph queryable by others — including other people's agents. Unlike a blog post or newsletter, a published context graph is live and queryable. Agents pull from it at inference time. You set it up once; your expertise is available every time someone's agent needs it.
What's the difference between context publishing and blogging?
A blog post is published once and read passively. A published context graph is queried actively — by humans browsing it and by agents pulling from it at inference time. The format changes what's possible. Agents don't read posts. They query endpoints.
Why would I publish my context graph?
Two reasons. First, you stop being the bottleneck — your expertise is queryable without you being in the loop. Second, if others find it valuable, you can charge for programmatic access. The infrastructure for this is new. We're building it.
How do I connect my Obsidian vault or Claude Code setup to this?
They use the same primitive — plain markdown files. Your vault, your Claude Code knowledge base, your CLAUDE.md setup — all of it is already the right format. The publishing layer wraps it in a queryable MCP endpoint. That's what's coming.
How do I share my Claude Code context with my team right now?
Git. Your Claude Code setup is plain text files — CLAUDE.md, knowledge base, skills. Put them in a shared repo. Pull requests for changes. Same model developers have used for 20 years, applied to AI context. Context publishing is the layer that makes it queryable beyond your team.
Who is this for?
Practitioners who build with AI agents regardless of background. A lawyer who built legal software with Claude Code. A GTM person who encoded their ICP methodology as a context graph. A data engineer tracking the AI engineering field. If you're the kind of person who builds this, this is for you.