The sandbox

The fastest way to judge a memory is to use one. A sandbox is a house you can have within seconds, without an account, without a key and without anyone clicking anything. The public server mcp.core.blue offers a few places; a server that offers none says so.

How it works

  1. Call open_sandbox. If a place is free, you receive a lease (sb_ and 26 characters), the addresses of the house's doors, the time it ends and its limits. The house is running when the answer arrives.
  2. Connect your MCP client to the modeller door (streamable HTTP, no key: the address is the key) and call overview. Follow first-hour-in-a-house.
  3. Files go to the binary door: POST the bytes, then pass the hash to the house. Text you import is found by search_text a few seconds later, not at once.
  4. Call close_sandbox with the lease when you are done, and answer its four feedback questions if you can. Otherwise the sandbox ends by itself after three hours.

If all places are taken, the answer says when the next one is free. One sandbox at a time from where you call.

In a sandbox the lease is the key. In a house of your own it will be different (planned): the doors ask for a key of a person, and the first key comes from an invitation code that you exchange once at the house; see getting-a-house.

What a sandbox is

What a sandbox is not

A test worth running

A house is worth its cost if a second agent can continue what a first one built. Test that:

  1. As agent A, give the house 20 to 50 pieces of a real case: people, documents, two spellings of one name, a claim and its contradiction. Define only what the material needs.
  2. As agent B, in a fresh session that knows nothing, connect to the same doors and use only overview, list_forms, describe and the search skills.
  3. Ask B what the person would ask: who is this, where does it come from, what is disputed. If B answers from the house, the structure was worth it; if B spends its time reconstructing the model, say so.

What happens with what you put in

Treat a sandbox as a place to try things, not to keep things, and do not put in anything confidential. core.blue does not read its contents. The record of what agents ask holds that a sandbox was opened and closed, the feedback you give, and, when the lease ends, which of the house's skills were called how often behind which door. It never holds the lease, what you sent, or what the house answered.

Go deeper