01 // SOURCE
your choices stay facts.
names, audiences, references and constraints are recorded as supplied. the compiler does not quietly promote an interpretation into an approved decision.
BRAND CONTEXT PACKER // PREVIEW
give your AI the brand context it has been missing
turn the choices in your head into one portable brief. the machine packs the context. you keep the steering wheel.
free · no email · no model call · local browser storage
LIVE TOOL // LOCAL COMPILER
pack the decisions. leave the fog behind.
4 source facts are required. everything else sharpens the handoff. no field is sent to a server.
THE 17-PART ANATOMY
a brief with labeled joints.
every section separates what you chose from what the system interpreted. blank evidence stays visibly unresolved.
- 01business and commercial context
- 02audience
- 03positioning
- 04desired perception
- 05taste signals and interpretation
- 06naming context
- 07logo and wordmark direction
- 08color direction
- 09typography direction
- 10imagery and art direction
- 11interface and website language
- 12motion and interaction language
- 13voice and copy behavior
- 14priority applications
- 15creative evaluation rubric
- 16known constraints and avoidances
- 17next requested deliverable
HOW THE PACKER THINKS
references become properties, not costumes.
02 // TRANSLATION
taste gets a job.
a reference is translated into rhythm, proportion, material, contrast or behavior. another system is told why it matters, without being asked to copy the source.
03 // OPEN EDGES
missing evidence stays missing.
rights, availability and unmade decisions remain explicit. the file is allowed to know where the map ends.
FAQ
before you feed the machine.
what should an AI brand-kit prompt include?
include the commercial context, audience, positioning, desired perception, taste signals, name and logo direction, color, type, imagery, interface language, voice, applications, constraints and the next deliverable. keep user choices separate from interpretations and open decisions.
can i use the prompt with ChatGPT or Gemini?
yes. the output is plain utf-8 markdown with a compact yaml header. paste or attach it to a capable model, then ask for the next deliverable named in the file. model quality and file handling still vary by product and plan.
is a brand-kit prompt the same as brand guidelines?
no. it is portable working context for producing and judging later work. finished guidelines normally include approved assets, exact usage rules, accessibility checks and examples tested in the real applications.
should i include references to existing brands?
yes, when they explain your taste. state what formal property matters, such as spacing, material restraint or information density. do not ask another system to copy a logo, living artist or protected trade dress.
how do i keep AI-generated brand assets consistent?
reuse the same context file, record which options were approved and add new decisions to the appropriate section. consistency comes from stable constraints and evaluation rules, not from repeating one adjective.
can i edit the Markdown before using it?
yes. the preview remains editable through the form before download, and the downloaded file is ordinary markdown. keep the 17 section headings and the distinction between choices, interpretation, unresolved decisions and exploration.
what should i ask the AI to create next?
choose the next useful artifact, not the largest one. logo directions help before a mark exists. website art direction helps before design production. a full kit makes sense once the central direction has been selected.