Lytics Content

The home for content work in Lytics. Today it covers content affinity; other content capabilities belong here as they are added.

Purpose

Help a Lytics customer get a topic map that actually means something, then turn it into Supertopics and audiences they can activate.

Topics are the hardest part of content affinity to get right:

The topic map Lytics produces is a starting point, not an answer. It depends on how pages were scraped, what the language-processing classifier (Google NLP by default) made of them, and whether the customer controls their own site markup.
Most customers never built an intentional taxonomy for their website, or for their product catalogue when a custom context layer sits on product data.
A topic set much larger than ~500 dilutes every score. Profiles whose top topics are too generic ("Home", "Customer") or too specific (one-off topics) are useless for activation.

This skill does three things, each in its own reference file:

Mode	What it does	Token needed	File
`assess`	Audits one or more existing context layers: content, topics, Supertopics, profile scores, downstream use. Optionally checks the live website. Produces a scorecard with evidence.	Read-only view token	`assessment.md`, `site-assessment.md`
`advise`	Turns the assessment into an advisement plan: a target taxonomy, proposed Supertopics, micro-audiences, and an ordered list of actions a person or an agent can carry out.	Read-only view token	`advisement.md`
`execute`	Carries out the approved `API` actions: topic and exact-page blocks, Supertopics it creates, and audience handoff. Writes step-by-step instructions for everything done in the UI or by a person (custom topic rules, page topic edits, re-classification, domain and path blocks).	Write token with only the scopes each action needs (see `execution.md`)	`execution.md`