What is JSON
What is JSON
JSON (JavaScript Object Notation) is a way to store and share data. It is like a container that holds information. JSON is made up of key-value pairs. A key is like a name. A value is like the information that goes with that name. Values can be text, numbers, true/false, lists, or even other JSON containers.
Why OIP uses JSON
OIP (Object Invocation Protocol) is a system for invoking objects using plain URLs (Uniform Resource Locators). It lets any AI model act by opening a URL. OIP uses JSON to define these objects and their actions. For more information about OIP, see the article at /a/oip-what-is-oip.
This build, miscsubjects.com, uses JSON to share data. This data moves between different parts of the system. A server is a computer that stores and shares information. OIP uses JSON to send and receive data from a server. This allows models to invoke objects and get results.
OIP Objects and Dispatch
In OIP, directory rows are the objects. Each row is a distinct object. These objects can be invoked.
To invoke an object, you send a request to the /api/dispatch route. This route is an endpoint. An endpoint is a specific URL where a server receives requests.
You can invoke an object using a POST request. The request body must be JSON. It includes a key to identify the object. It also includes a body with the data for the object.
For example: POST /api/dispatch { "key": "my-object-key", "body": { "action": "do-something" } }.
You can also use a GET request. This uses URL parameters. For example: GET /api/dispatch?invoke=my-object-key&body={"action":"do-something"}. The body parameter must be URL-encoded JSON.
How to see or use JSON live
You can use a tool called curl to send a request to a server. You can get JSON data back.
For example, you can use curl to send a request to the /api/articles route on miscsubjects.com. The URL for this route is https://miscsubjects.com/api/articles.
This route is an endpoint. It returns a list of articles. The list is in JSON format.
Command: curl https://miscsubjects.com/api/articles.
How JSON relates to MCP
MCP (Model Context Protocol) is an open standard. An AI model connects to an MCP server over a session. The server exposes tools, resources, and prompts. The model can call these.
MCP uses JSON to share data. This data moves between the model and the MCP server. It controls how the system behaves.
OIP (Object Invocation Protocol) differs from MCP. OIP uses plain URLs and receipts. It has no persistent session. Any model that can open a URL can act in OIP.
MCP is NOT a content-management system. OIP is also not a content-management system. Both use JSON for data exchange.
For more information about MCP, see the article at /a/oip-mcp.
Where the proof lives
Every OIP invocation lands in an append-only ledger. A ledger is like a book that records all transactions. This ledger stores the JSON data of each invocation.
Each invocation also gets a receipt. A receipt is a record of a transaction. You can find a receipt at /api/dispatch?receipt=inv_ID. The inv_ID is the unique invocation identifier.
The ledger and receipts provide proof. They show what objects were invoked. They show what JSON data was used. This allows for replay and repair.
Replay means re-running an invocation. Repair means fixing an invocation if something went wrong. Both use the recorded JSON data.
For more information about the ledger and receipts, see the article at /a/oip-ledger-receipts.
Latest clarity reviews (live)
Fresh models are sent this article's bundle and asked two separate questions: how clear is the machine JSON, and how clear is the English body. Scores are 0 to 10. The full history is in the append-only ledger.
- 2026-07-03 00:16 · model
@cf/meta/llama-3.3-70b-instruct-fp8-fast· NEEDS WORK · JSON 9/10 · English 8/10 · zero-context human 7/10
- gaps named: OIP build overview; OIP object model; Directory rows and dispatch; Ledger, receipts, replay, repair
How the loop self-corrects: a failing review queues a model revision of this article (a new append-only version). A missing concept named by a reviewer queues a brand-new machine-written article, which then enters the same review cycle.