brand.json is a machine-readable, public projection of a brand's record, built for other agents to read. It is a deliberately bounded slice: enough for an agent to represent the brand faithfully, with the proprietary strategy held back. A public face for the record, not the whole record.
Why it matters
Agents are starting to act for brands and on brands: answering about them, buying from them, representing them. Each one needs a reliable place to read who the brand is, and a scraped web page is a poor and forgeable substitute.
A published projection fixes that: a stable, machine-readable statement of identity, voice, and positioning an agent can read directly instead of inferring. It is the brand speaking in the format agents actually consume.
The catch is what you expose. A public brand file should carry enough to be useful and nothing that hands a competitor your playbook, and that line has to be drawn deliberately, not left to a default.
How Jinn treats it
Jinn's public projection is an explicit allowlist, not a filtered dump. A curated set of identity, voice, and positioning attributes is published; everything else, competitive intel, pricing, product internals, the record's own metadata, is withheld by simply not being on the list.
Adding anything to that public face is a deliberate act, and the code is written so a new internal attribute leaks nothing by default. The public brand file shows the brand is grounded without giving the strategy away.
Related terms
A public face for the record, drawn deliberately.
Enough for an agent to represent you faithfully, with the strategy held back. See how the record is built.