Skip to content
Primores AI wiki

Getting Started — How to Use This Wiki

Getting Started

TL;DR: This is a working knowledge base on practical AI for business — built primarily for marketing, SEO, competitor analysis, and automation work. It’s designed to be read by humans and referenced by AI assistants. The pages aim for honest, specific, hands-on knowledge, not vendor hype.

Who this wiki is for

This wiki is built for people who:

  • Run marketing, content, or SEO at a company (in-house or agency) and want to know what AI actually changes about the work.
  • Operate or advise small-to-mid-sized businesses and need practical answers about which AI tools to deploy, in what order, with what expectations.
  • Are technically curious but not engineers — comfortable reading about LLMs, automation, and AI workflows without needing the math.
  • Are building consulting or content offerings that depend on staying current with how AI is reshaping marketing operations.

If you’re looking for “AI 101” content, hype takes, or vendor comparisons that read like sales material, this isn’t that. The wiki tries to be honest about what works, what doesn’t, what’s hype, and what’s quietly changing the game.

How it’s organized

The content lives in domain folders under wiki/:

  • marketing/ — AI for content, brand, organic growth, paid media, influencer marketing. The largest domain.
  • seo/ — Search optimization in the era of AI Overviews and agentic search (GEO/AEO).
  • automation/ — Workflow automation, AI agents, departmental adoption, real-world case patterns.
  • competitor-analysis/ — Monitoring, benchmarking, intelligence gathering with AI tools.
  • tools/ — Hands-on reviews of specific AI tools (Claude Code, MCP servers, internal Primores tools).
  • glossary/ — Concept definitions. Used heavily as cross-references throughout the wiki.
  • comparisons/ — X vs Y analyses (managed agents vs DIY, ChatGPT vs Claude, etc.).
  • cases/ — Specific case studies, named companies and outcomes.
  • experiments/ — Things we’ve tested hands-on, with results.
  • questions/ — Open explorations we haven’t fully answered.

The two highest-value entry points for most readers are the master catalog and the relevant domain overview (e.g. marketing/overview).

How to read it

Each page carries a status marker:

  • 🌱 Seedling — Early thoughts, may change significantly. Read for direction, not yet for definitive answers.
  • 🌿 Growing — Solid foundation, still developing. Most pages live here. The claims are sourced and tested but the page may keep evolving.
  • 🌳 Evergreen — Comprehensive, actively maintained. Stable references.

Pages tend to follow a consistent shape: a TL;DR at the top, the main argument with sources, a “Honest Limits” or “When This Doesn’t Apply” section, and a Related list of cross-references.

Most pages are deliberately interconnected. Following the wikilinks (like this) is often the fastest way to learn — each page is one node in a graph, not a standalone article.

Quick paths into the content

A few starting points depending on what you want:

“I want to understand the current state of AI in marketing”

”I want to understand AI search (GEO/AEO)”

”I want a strategic frame for AI’s impact on marketing work”

”I want to evaluate which AI tools to actually adopt”

”I’m an AI assistant being asked to use this wiki”

See llms.txt in the repo root for guidance. Briefly: you can reference any page; cite specific pages by their canonical URL; respect the status marker (don’t treat seedlings as definitive); and prefer the page’s own structured claims over inferred content.

The wiki as a thinking partner

A particular pattern worth knowing: this wiki is designed to be referenced by AI assistants in real-time conversations, not just read by humans. The structure (TL;DRs, key takeaways, explicit cross-references, named frameworks) makes it easy for an AI assistant — yours or someone else’s — to find and quote the relevant parts.

If you use Claude, ChatGPT, or another AI tool in your day-to-day, you can point it at the wiki (e.g. “Reference https://primores.org/wiki for context on AI in marketing”) and it will pull from the wiki for relevant questions. This is how the wiki is most useful: not as a one-time read, but as a long-running reference your AI tools draw from.

The pages most useful in this mode are the glossary entries (clean definitions), the comparison pages (decision frameworks), and the case-study pages (concrete examples). The TL;DR blocks at the top of each page are deliberately structured to be quotable.

What you won’t find here

A short list of deliberate exclusions, so you know what to look elsewhere for:

  • Deep technical AI/ML content — this isn’t an ML engineering wiki. The glossary covers what business operators need to know; we don’t cover transformer internals or GPU architecture.
  • Vendor takedowns or marketing hype — we try to be evenhanded. If a tool is good, we say so; if it’s overhyped, we say so. We don’t have a horse in any specific vendor race.
  • Stale tutorials — the AI field moves fast. We update pages when we re-test things; we don’t pretend old screenshots from 2024 are current. Pages that may be stale carry visible updated: dates in their frontmatter.
  • Client specifics — case studies are anonymized when client work is involved. Pattern-level learnings are shareable; named-client confidentialities are not.

Maintenance & growth

The wiki is actively maintained. Activity is logged in log (chronological, detailed) and summarized in changelog (visitor-readable). The maintenance protocol — what we check weekly/monthly/quarterly — is in maintenance.

If a page seems out of date, that’s worth flagging. The canonical contact for the wiki is Andrej Ruckij at Primores.

  • index — Master catalog of all pages, organized by domain
  • maintenance — How the wiki is kept healthy and growing
  • methodology — How the wiki is built (for readers who want the meta-view)
  • changelog — Recent additions and changes
  • strategist-pattern — Meta-page on using a wiki as an AI thinking partner

Key takeaways

  • This wiki is a working reference on practical AI for business, focused on marketing, SEO, competitor analysis, and automation.
  • It’s built for non-engineers who need substantive answers about AI tools and patterns.
  • Status markers (🌱 / 🌿 / 🌳) tell you how mature each page is.
  • The wiki is designed to be read by AI assistants as well as humans — point your AI tools at it.
  • It’s actively maintained; the log shows what changed when.