Collection+JSON is a JSON-based hypermedia media type for managing and querying collections of resources. It defines a standard document structure for items, queries, templates, and links to support read and write operations.
Collection+JSON
Collection+JSON, registered as the media type application/vnd.collection+json, is a JSON-based hypermedia format created by Mike Amundsen in 2011 for managing and querying collections of resources. Rather than returning bare JSON data, it wraps items in a document that also carries the links, queries, and write templates a client needs to navigate and change the collection. It was one of the earliest attempts to make plain JSON APIs genuinely hypermedia-driven.
- A fixed document structure -
collection,items,links,queries, andtemplatesections that every client can parse the same way. - Write templates - A
templateobject tells a client exactly how to construct an item to add or update, so the server drives the write. - Queries as data - Supported searches are described in the response itself instead of being baked into out-of-band documentation.
- Read and write in one media type - The same format covers navigating a collection and mutating it.
Collection+JSON showed up in the hypermedia conversation alongside formats like HAL, Siren, and JSON-LD, and it remains a clean teaching example of why relationships and actions belong in the response, not just the data. In practice most APIs never adopted a full hypermedia media type, but the ideas it pushed — self-describing links, server-supplied write templates, and machine-discoverable queries — are exactly the affordances that matter again now that autonomous agents, not just humans, are the ones reading responses and deciding what to do next.
Referenced on the API Evangelist blog
Where this standard shows up across sixteen years of my writing at apievangelist.com — how it fits into API design, governance, and the agentic turn.
