Brand Kit — Context Layer

Do you need one?

Only if you generate content. If your agent ingests sources and queries them, skip this page entirely — nothing in step 1 requires a brand kit.

Without one, the content engine has no identity to work from and output reads generic.

The six fields

guidelines accepts these and nothing else. An unknown key returns 400.

brand_name

Appears in headings, references and attribution. Without it the engine uses placeholders.

brand_domain

Internal links and CTAs point here. Also what makes a citation count as yours.

brand_description

What the organization does. A strong one produces more on-topic content.

voice_and_tone

The highest-impact field. Specific beats vague by a wide margin.

author_persona

Expertise and perspective. A developer advocate writes unlike a financial advisor.

global_writing_rules

Hard rules on every piece of output, layered with content-type rules.

bash

curl -X PUT https://apiv2.senso.ai/api/v1/org/brand-kit \
  -H "X-API-Key: $SENSO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "guidelines": {
      "brand_name": "Acme Credit Union",
      "brand_domain": "acmecu.com",
      "brand_description": "Member-owned credit union serving small businesses since 1985",
      "voice_and_tone": "Warm, knowledgeable and jargon-free. Like a trusted neighbour who happens to be a financial expert.",
      "author_persona": "Senior financial advisor with 15 years of experience",
      "global_writing_rules": [
        "Use active voice",
        "Keep sentences under 25 words",
        "Avoid acronyms unless defined on first use"
      ]
    }
  }'

Voice and tone is where the quality is

Specific beats vague, and it is not close.

Vague: “Professional and friendly.” Produces forgettable content.

Specific: “Warm, knowledgeable and jargon-free. Like a trusted neighbour who happens to be a financial expert. Explain concepts as if the reader is smart but unfamiliar. Use analogies from everyday life.”

How to get this out of someone

Ask: who do you want to sound like when talking to your customers? Their answer maps straight onto this field.

Write rules a reviewer can check

Good rules are binary. Someone can answer yes or no without arguing.

Write clearly

Keep sentences under 25 words

Be concise

Articles must be under 800 words

Use simple language

Write at an 8th-grade reading level

Be consistent

Avoid acronyms unless defined on first use

Updating it

PATCH merges the keys you send. PUT replaces the whole guidelines object, so read, merge, then write, or you will drop fields you meant to keep.

bash

# PATCH merges only the keys you send. PUT replaces the whole object.
curl -X PATCH https://apiv2.senso.ai/api/v1/org/brand-kit \
  -H "X-API-Key: $SENSO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"guidelines": {"voice_and_tone": "Casual and direct"}}'

An empty patch is an error

{"guidelines": {}} returns 400 — a patch must include at least one valid field.

bash

senso brand-kit get --output json
senso brand-kit patch --data '{"guidelines": {"voice_and_tone": "Casual and direct"}}'
senso brand-kit set   --data '{"guidelines": { ... }}'   # full replacement

How it reaches the output

The brand kit is org-wide; a content type is per-format. Both sets of writing rules apply during generation: global rules for organization-wide standards, content-type rules for format-specific ones.

Brand Kit — voice, persona, global rules. How it sounds.
Content Type — template and format rules. How it is shaped.
Knowledge Base — the facts. What it can say.

Errors

400 — unknown key in guidelines, wrong type, or an empty PATCH
401 — missing or invalid API key

Knowledge Base Build the Context Layer API reference

Next section