Your playbook of preferences, skills and project context. Every AI app you use starts with it and keeps it current.
Free for up to 3 AI apps.
and any app that supports MCP
Both of these happened to us while we were building Context Locker.
We wrote our product plan in a Claude doc. Codex can't open it, so it can't help build the product unless we paste the plan in and keep two copies in step.
With Context Locker, the plan is a note in the project, and every AI you connect can read it.
We'd tuned our web design skills in Codex. To use them with Claude, we had to go back to Codex, export them and hand them over through GitHub. Every time we improve one, we have to do it all again.
With Context Locker, skills live in one place with their source and every version. Every AI uses the same ones, with nothing to export.
Memories, rules, skills and docs are all stuck like this today. Use more than one AI and you either repeat yourself or lose the work.
Pulls in what your AIs already know, merging repeats and flagging clashes.
Every AI you connect starts each session with your latest playbook.
AIs suggest updates as they learn. Approve once and every AI has it.
Models, tools and the way you work keep changing, so skills need regular updates. Instead of updating a copy in every app, keep one in Context Locker that every AI uses.
Release notes skill
In your playbook, used by 3 AI apps
v4 by Codex, approved by Mary on 14 Oct 2026
What changed
v4 is in use in all 3 apps. Pick an older version to see what changed, or revert to it.
Every change shows who made it and why. Roll back in one click.
A monthly check-up flags unused skills, clashing rules and broken links.
Found a better skill? Compare, then swap it in everywhere. The old one is archived.
Share a skill once and every AI on your team uses the same version. Coming with team plans.
Plans, specs and decisions end up wherever an AI happened to write them. Notes gives each project one home for its docs, and every AI you connect can read and add to it.
Docs in a repo or Notion stay where they are. If a project has none, your AI asks once and Notes becomes its home.
Every note has a project, tags and a one-line summary, so your AIs search and read only what a task needs.
Your AIs write notes freely. Each change keeps who made it and a short note on why.
Open your notes in Obsidian or any editor, and move them out with one command.
It doesn't know how you work, what you care about or what you're building. So you teach it. You explain your rules, correct its habits, write down the steps that work and brief it on every project. Slowly, it becomes useful.
That teaching is real work. It's a layer you build on top of the model, and it's the part that's about you.
Today that layer is scattered across apps and locked inside each one. Change assistants, or try a better model, and you start again from zero.
We think the layer belongs to you. You should own it, keep it current and take it anywhere: to any model, any assistant, any agent.
Context Locker is where it lives. One playbook of your preferences, skills and project context, in plain files you can read, kept sharp by the AIs that use it.
Models will come and go. What you've taught them should stay yours.
Export your playbook and notes as Markdown and skill folders, any time.
Built on Agent Skills, AGENTS.md and MCP, the formats your tools already use.
Your AI writes plain files and a one-line refresh. No extra app runs on your computer.
Encrypted in transit and at rest, never used for training. You choose which AI apps can read it.
$0
For trying it with a few AI apps.
$15a month
Or $150 a year: 12 months for the price of 10.
Bringing your team? Team plans are on the way. Join the waitlist
Their memory lives inside one app, and you can't take it with you. Your playbook works in every AI you connect, you approve what goes in, and you can export it as plain files.
Yes. Claude Code, Codex, Grok Build, Hermes and OpenClaw get your playbook in the files they already read. It refreshes when each session starts, or once a day for always-on agents like OpenClaw and Hermes.
No. Your playbook covers anything you hand to AI: writing, research, analysis, planning and more. Coding agents simply get it in their own files too.
No. If a project's docs already live in a repo or Notion that all your AIs can reach, Context Locker points to them. Notes is for projects without a home like that, and you can move notes out any time.
No. ChatGPT, Claude, Grok, dots and Muse connect over the web. Coding agents like Claude Code and Codex read their instructions from files on your computer, so you ask your agent to set up Context Locker. It downloads your playbook through our API, writes those files and sets up a one-line refresh that runs when each session starts, or once a day for always-on agents like OpenClaw. No extra app runs on your computer.
It's encrypted in transit and at rest, and never used for training. You choose which AI apps can read it. Export everything, notes included, as Markdown and skill folders whenever you like.
Free for up to 3 AI apps. Set up by asking your AI.